LARSHAMMERSTRÖM MechanismsandPhenomenainBrakingandGripping 199 DigitalComprehensiveSummariesofUppsalaDissertationsfromtheFacultyofScienceandTechnology

Full text

(1)Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology 199. Mechanisms and Phenomena in Braking and Gripping LARS HAMMERSTRÖM. ACTA UNIVERSITATIS UPSALIENSIS UPPSALA 2006. ISSN 1651-6214 ISBN 91-554-6597-8 urn:nbn:se:uu:diva-6974.

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10) ! " # $ $ $ % & '

(11)

(12) (

(13)

(14) (& ) *& !& +

(15)

(16) %

(17)

(18)

(19) , -

(20)

(21) .

(22) & /.

(23)

(24) &

(25)

(26)

(27)

(28) "00& 10 & & 2,3 0"45514!50647& /

(29)

(30)

(31) $

(32)

(33) $ $

(34) 8

(35)

(36) 8

(37) & '

(38) $

(39)

(40) $

(41) $

(42)

(43)

(44) $ (

(45) $ & /.

(46)

(47)

(48)

(49) $

(50)

(51)

(52)

(53) - ( ( $

(54)

(55) & ' $ $

(56) $ -

(57) $

(58) & ,

(59)

(60)

(61)

(62) (

(63)

(64) -

(65) $

(66) $

(67)

(68)

(69) 9

(70) $

(71)

(72) & ' $

(73) . "

(74) -& 2 $

(75)

(76)

(77)

(78) & ' -

(79) (

(80) $

(81) - (

(82) & 2

(83) ( 1 .:;<& =

(84)

(85) $ -& '

(86) ( - (

(87) .

(88)

(89) = $$

(90) & ' $

(91)

(92) ( (

(93) & ' $$

(94) ( $

(95)

(96) $

(97)

(98) $

(99) . $ & 2

(100)

(101)

(102) = (

(103) 8 .

(104) $

(105)

(106)

(107) $

(108) & '

(109) $

(110)

(111) $

(112) $

(113)

(114) $

(115)

(116) $$

(117) $ $

(118) & 2 ( (

(119) $ ( - $

(120) & / ( $

(121) . ( $$

(122) $ $

(123) $ & '

(124)

(125) $ $ ( $

(126)

(127) $ $

(128) $ 4$

(129) 8

(130) &

(131) , - ' = !

(132)

(133) "

(134) #

(135) $ # % & '()# # $*+',-, # > * ) ! 23 "!5"4!"1 2,3 0"45514!50647

(136) #

(137)

(138) ### 4!061 ? #;;

(139) &-&; @

(140) A

(141) #

(142)

(143) ### 4!061B.

(144) To my beloved family.

(145)

(146) List of papers. I. II. III. IV. V. VI. Friction flims on brake pads -Tribologically induced formation of nanocrystalline layers L. Hammerström, E. Coronel and S. Jacobson In manuscript Surface modifications of brake discs for fundamental studies of the generation of brake squeal L. Hammerström, M. Hanson and S. Jacobson In manuscript Surface modification of brake discs to reduce squeal problems L. Hammerström and S. Jacobson Wear, 261, (2006), 53-57 Brake squeal reduction by particle embedding in the disc surface –influence of treated pattern L. Hammerström and S. Jacobson In manuscript Pressure sensitive film as a tool for investigating the pressure distribution on brake pads L. Hammerström and S. Jacobson In manuscript Designed high-friction surfaces –influence of roughness and deformation of the counter surface Hammerström and S. Jacobson In manuscript. Published papers are reproduced with permission from the publishers..

(147) The author’s contribution. The author of this thesis has performed the major part of planning, experimental work, evaluation and writing for all included papers excluding the TEM analysis in paper I..

(148) Contents. Introduction.....................................................................................................9 Tribology and friction ................................................................................9 High friction applications.........................................................................10 This thesis.................................................................................................11 Equipment ................................................................................................11 Flat-on-flat rig......................................................................................11 Brake rig ..............................................................................................12 FIB .......................................................................................................12 High-friction applications .............................................................................14 Braking and gripping materials ................................................................14 Brakes and friction...............................................................................14 Textured brake discs ............................................................................21 Gripping surface (Paper VI) ................................................................33 Summary in Swedish ....................................................................................40 Bromsar ...............................................................................................41 Greppytor.............................................................................................44 Acknowledgements.......................................................................................47 References.....................................................................................................48.

(149)

(150) Introduction. Tribology and friction The continuous technical development to meet the demands of the society today requires a deeper knowledge of phenomena related to surface interaction. The driving force of these ever increasing demands on improving the tribological properties of materials is of course lower cost and longer life of products, but also the concerns for our environment. Ultimately the goal is to be able to design materials and surfaces to give the best function, the best economy and the least environmental stress possible. By controlling the coefficient of friction, a great step is taken towards a more energy efficient world. When looking closely onto something it often occurs that things are not as they seem. This applies very well to surfaces of materials. You look at what you think is a piece of metal, say a coin, but in reality you are looking at metal covered by an oxide and then by adhered water, fat and salts from human fingers, bacteria and all kind of dirt. So, the first thing a material scientist has to realize is that the surface of a material in most cases is very different from the bulk. Then; what is friction? Friction is the reaction force restricting the relative motion of two contacting surfaces [1]. The coefficient of friction is given by the formula:. µ. Friction force Normal force. One has to understand that the coefficient of friction is not a material property. It depends on the material properties such as hardness, melting temperature, Young’s modulus, etc, but also of geometrical parameters (roughness) and environmental parameters (temperature, atmosphere, lubrication). When rubbing two bodies, made of different materials, a third type of material with properties different from the two original materials will form a so called third-body layer [2]. This layer consists of wear particles blended and alloyed together from the original materials. Often this tribological situation occurs in air, which will oxidize the wear debris. One thing is for sure; the 9.

(151) mating surfaces are not the same materials as those in the bulk of each body. It is clear that due to this third-body layer, the same material could have vastly different coefficient of friction values in different situations.. High friction applications To achieve an optimized function of a machine, it is important to have the right friction level between every connecting detail. This often means low friction, but also high friction could be required. Applications relying on high static friction include various types of fixtures, couplings, bolted joints, torsion joints, etc. The common characteristic of these applications is that they rely on the friction force to maintain the relative position of two mating surfaces. Applications relying on high dynamic friction are also common, the main example being brakes where a low friction could be devastating. There are two ways to achieve a high coefficient of friction in a contact, either by applying a high load or by using surfaces with high-friction properties. To be able to apply a high load without suffering to much deformation, the construction needs to be very stiff and use a larger number of bolts which results in a heavy and expensive construction. A better solution is often to use a high-friction surface. Generally a brake would work better with higher coefficient of friction. It would be smaller and require less pressure on the pads. This would allow a design that could save material and energy, but there are other limitations to take into account when designing a brake system. It has to be cost efficient, which result in that only very cheap materials are candidates for both disc and lining material. Further, low wear rate is mandatory. The brake has to be reliable, i.e. show a stable coefficient of friction irrespective of outdoor temperature, humidity, disc oxidation, temperature gradients, etc. Also comfort is an important property. The brake has to give the right feeling and response to the driver and avoid unwanted side effects like noise and lately, not one day to early, an environmental friendliness is demanded. High-friction gripping surfaces are used in applications where no play in the joints is allowed, such as in assembly robots. It is a clever way to reduce the weight to use high-friction surfaces in the joints, since the joint pressure and thus the number of bolts can be reduced. Lower weight is well correlated to lower energy consumption and faster acceleration of moving parts.. 10.

(152) This thesis The aim of this thesis is to gain knowledge of how high-friction surfaces work, and increase the understanding of phenomena related to high friction. For the brakes the focus is friction related noise, i.e. brake squeal, and how to reduce it. For gripping surfaces the emphasis is put on the influence of counter surface roughness on the coefficient of friction. The outline is as follows: x An overview of the experimental equipment x The tribology of disc brakes x The tribology of textured gripping surfaces x Summary. Equipment Flat-on-flat rig The specimen holders of the flat-on-flat rig [3,4] allows the surfaces in contact to self align, see Fig. 1. By applying an increasing tangential force and continuously monitoring the friction, the force where the surfaces start sliding is found. The lower specimen is movable and the position and friction force is recorded on a computer. The upper specimen size is 4x4x6 mm and the lower specimen is 6x6x40 mm.. Figure 1. Schematic of the self aligning test samples in the flat-on-flat rig. The upper specimen can rotate in the sliding direction and the lower specimen can rotate transverse the sliding direction.. 11.

(153) Brake rig One of the main experimental equipments used is a brake rig. It is based on a brake from a Volvo 850, where a front left corner including spring strut and lower wishbone is connected to an electric engine via a gear box, see Fig. 2. The speed is adjustable between 0 and 4 revolutions per second, and the brake-line pressure between 0 and 30 Bar. Temperatures at the disc surface and inside the calliper are measured as well as the torque, brake-line pressure and sound pressure. When the brake is applied and reaches a brake-line pressure above 1.5 bar, a snap shot of the sound is captured every third second until the pressure is released. The sound snapshot is analyzed with fast Fourier transforms, and the main frequency and the sound pressure are saved onto a computer together with all the other measured data. A thorough description of this equipment is given in [5].. Figure 2. The brake squeal test rig. a) Schematic, and b) Photo of the test rig.. FIB Ions can be accelerated in an electric field. In the combined Focused Ion Beam / Scanning Electron Microscope (FIB-SEM) the ion beam can be used to sputter material away from the sample surface [6]. By doing so, a well defined pit can be carved out of the sample, allowing a cross-section image to be taken at any chosen point, see Fig. 3.. 12.

(154) Figure 3. Schematic of the FIB. The ion and electron beams are tilted 52º in respect to one another. The ion beam can be used both for removing material and for imaging the surface. Before preparation of a TEM-sample, a protective layer of platinum is deposited on top of the specimen.. 13.

(155) High-friction applications. Braking and gripping materials A prerequisite for a material used in applications needing high friction is the ability to produce atomic bonds to the counter surface. These bonds should, to avoid wear of the connecting materials, be weaker than the internal bonds of the material. That applies for the adhesive component of the friction. Most often the friction is dependant of the surface geometry in the sense that the roughness restrains the relative movement of the material. The latter is often called the ploughing component of friction as the harder surface scratches the softer surface [7].. Brakes and friction The friction couple of disc brakes comprise a disc, usually made out of grey iron, and organic or semi metallic brake pads. The microstructure of the grey iron disc is perlitic with 3-4 wt% carbon. Some of the carbon is found as graphite in the shape of small flakes which are embedded in the iron matrix and are essential for a good thermoshock resistance [8] There are several different types of brake pads, but in all experiments presented in this thesis organic pads are used. An approximate composition of a typical organic brake pad is given in Table 1. The ingredients can be divided into groups based on the properties added to the pad. These groups are presented here below [9]. x The structural materials which add mechanical strength to the material have shifted totally from mainly asbestos as structural fibre, to a complex mixture of fibres such as fibres of steel, brass, glass and kevlar. x Fillers with high temperature resistance and low price are used to spread out the fibres and improve manufacturability. Clay minerals like vermiculite and barium sulphate are commonly used.. 14.

(156) x As binder material different kinds of phenolic resin is used. These are relatively thermo-resistant, but will smear out at the pad surface or even evaporate during high temperature brakings. x To ensure a high and stable coefficient of friction, frictional additives are added. They are supposed to give some lubrication between the metallic fibres and the disc. Ceramic particles are added for removal of unwanted surface films such as corrosion products on the disc.. Table 1 Structural component Fibers Matrix Friction modifiers. Filler. Ingredient Steel, aramid and glass Binder Other Metallic (brass, bronze, iron) Graphite Metal sulphide Quartz Clay minerals, iron oxide. Amount [wt.%] 30 8 11 15 15 8 5 8. Experimental The pads used in the experiments are slightly modified standard brake pads to a Volvo 850. The modification has made them very prone to make squealing. No anti-squeal shims are used. Under these conditions these brakes normally squeal in 65-75% of all brakings and give an average coefficient of friction of 0.50 – 0.55. The brake sequence used is 42 brakings long, with a combination of different brake line pressures [10]. The brake sequence is repeated in sets of five sequences to a total run time of nine hours. Each disc is 290 mm in diameter and the width of the friction surface is 55 mm. Squeal Squeal is caused by uncontrolled self amplifying vibrations in different parts of the brake system. It could be the disc, the pads or the calliper that vibrates but the energy to drive the squeal always comes from the sliding frictional contact between pads and disc [11]. The squeal itself does not affect the performance of the brake, but it could indicate that a compositional change has occurred at the pad surface. A number of measures are normally taken to inhibit brake squeal [12]. The components of the brake are carefully designed to minimize the risk of having easily activated resonance frequencies. An important part of develop15.

(157) ing any new pad material is to give it a low squealing tendency. Further, damping shims are mounted at the backplate of the pad. The shims are developed to reduce the vibrations transferred from the pad to the calliper. The plateau model The real contact between the disc and the pad arises in small micro sized contact spots spread out over the most protruding parts which are called primary plateaus, see Fig. 4 [13]. Between these contact spots there are gaps large enough for the wear debris from both disc and pad to easily fit in and make way. Occasionally a wear particle hit a contact spot and is hindered in its way through the contact between the pad and the disc. First, only a small fraction of the larger wear particles get trapped, but as some particles get trapped more and more particles join the group, forming the seed of a secondary plateau. The plateau growth continues making the space between the growing secondary plateau and the disc smaller, trapping particles in even greater number and smaller sizes. This far, the smallest nano-sized particles have been virtually unaware of the debris compacting going on, but as a final step even the nano-sized particles get caught in the contact. When rubbing a pad against a disc made of glass, the entrapment of the nanoparticles actually gives an optical phenomenon where it looks like wave fronts travelling over the plateau in the opposite direction of the disc sliding direction [14]. The secondary plateau eventually finds a steady state when the degrading processes and the compaction of new particles balance for the prevailing circumstances. However, a small change of the contact situation can disturb the balance and the plateau will brake lose only to be milled down into wear debris again. The wear debris will either be reused in another secondary plateau or finally exit the contact. Most of the iron oxide as well as the fragments of iron originate from the grey iron disc [15] and are either oxidized before removal or oxidized during milling of iron fragments in the interface between pad and disc.. 16.

(158) Figure 4. Schematic of the contact situation between disc and pad according to the plateau model. The disc is symbolized by a transparent plane and is sliding from left to right. Protruding constituents, i.e. primary plateaus of the pad are white, compacted debris in the form of secondary plateaus are grey. A constant flow of wear debris in the gap between pad and disc wear the lowlands of the pad through three body abrasion and supply the secondary plateaus with new material. Occasionally secondary plateaus break down, releasing heavily deformed particles back to the flow of wear debris.. Features of secondary plateaus (Paper I) One way to learn more about the pad-disc contact and the build up of a secondary plateau is to study TEM-samples from the interface between a primary and a secondary plateau. In Paper I two TEM-foils are studied. In one sample the primary plateau is based on a bronze flake and in the other on a steel fibre. The pad is of the organic type described earlier. It was previously shown that secondary plateaus are created during use and that a hard layer takes shape on top of the plateaus [13,16]. Nanoindentation on a secondary plateau down to 50 nm gave a hardness value close to 4 GPa. This value decreased rapidly for deeper indentations giving 0.8 GPa and 0.2 GPa for 400 nm and 1 µm depth respectively. This hard surface layer was thought to consist of particles smaller than 10 nm. The secondary plateau comprises mainly iron oxide but also iron fragments in varying sizes and other fragments originating from the pad. An indication of the high stress conditions the iron particles are subjected to is the shape of some of the superficial iron particles in Fig. 5. The particles are deformed and stretched out by the tremendous material flow caused by contact against the disc. The superficial material flow will be discussed later.. 17.

(159) Figure 5. The bronze TEM-sample from a brake pad, analyzed in TEM bright field mode showing mainly diffraction contrast. The counter surface sliding direction has been from the left and thus compacting the debris against the bronze at the right hand side of the image. The areas marked with rectangles In the TEM image represent the image areas of Fig. 6 and Fig. 7. A schematic of the bronze TEM foil where important features are pointed out is presented below the TEM-image.. Also the primary plateaus are affected by the force from the disc. An expansion of the crevice approximately 3 µm below the surface reveals that the secondary plateau actually has been displaced 0.5 µm to the right. Two areas of transferred bronze confirm the plateau movement, see Fig. 6. Simultaneously with the plateau displacement, two voids have appeared in the tin bronze flake. The formation of the secondary plateau and the subsequent creation of these voids explain why the voids are free from iron oxide debris. Altogether, this plateau displacement shows one deteriorating mechanism where the primary plateau has been sheared. If the process continues the secondary plateau will lose the support from the primary plateau and the iron oxide will be milled down into debris again.. 18.

(160) A. C D. D. B A. Figure 6. A close up of the interface between the bronze and the secondary plateau reveal many interesting features. The interface between the primary (A) and the secondary (B) plateaus show a very sharp line with no trace of alloying, and the secondary plateau seems very fine grained except for some larger iron rich particles (C). It looks like the compacted material does not fill all of the cavities of the bronze even though the cavity is huge compared to the grain size of the milled debris. The arrows marked D point out material that has been sheared from the bronze as the secondary plateau has been pushed forward by the friction force from the disc contact. The cavities have probably been formed during the same process, which explains why they are void instead of filled by iron oxide debris.. From the surface and down to a depth of about 50 to 100 nm, the iron oxide is a mixture of nanocrystalline and amorphous particles, revealing that a heavy deformation has taken place in that region. A line, sometimes widened to a narrow crevice, separates the deformed surface zone from the nanocrystalline material below, see Fig. 7. A schematic of the material flow in the friction film is shown in Fig. 8.. 19.

(161) D A B. C. Figure 7. The semi-amorphous frictionfilm (A) and the nanocrystalline friction layer (B) divided by a horizontal border (C). On top of the deformed zone the deposited platinum appears as white particles (D). There is a clear difference in grain size between the material above and the material below the border. Right below the border, grain sizes up to 10 µm are common, while 5 µm sized particles dominate in the deformation zone where even amorphous areas are seen.. Figure 8. Schematic of the material flow of the friction film. A flow of nanoparticles is created by the sliding disc contact. The particles closest to the disc have a higher speed than particles deeper in the friction film. The particles in the friction layer are blending while the material below moves along the turbulent nanoparticle flow.. 20.

(162) According to computer simulations [17, 18], a turbulent deformation layer of a quasi-liquid is likely to form between two metallic surfaces during a sliding motion. This turbulent deformation layer will have a boundary approximately 50 particle units below the surface where the material flow will become lamellar. Assuming that the average size of a particle in the turbulent zone is 2 nm the layer would be around 100 nm thick, which is in accordance with the situation seen in the TEM-images. The TEM study of the secondary plateau has confirmed the presence of a hard frictionfilm and found relevance in computer simulations of the behaviour and boundary of the friction film.. Textured brake discs An effect connected to the formation the secondary plateaus is the increasing coefficient of friction during each individual braking [19], see Fig. 9.. Coefficient of friction. 0,7. Last measurement. One braking. 0,65 0,6 0,55 0,5 0,45 0,4 200. First measurement. 201. 202 203 Number of brakings. 204. 205. Figure 9. The increase in coefficient of friction within every single braking of an untreated brake disc against TX4005B brake pads. Within each braking nine measurements are made. The increase in coefficient of friction is about 0.1 to 0.2 within one 20 seconds long braking with constant brake line pressure.. The switch of materials in contact, from a mixture of different pad constituents versus iron oxide to mainly iron oxide versus iron oxide could, as well as the higher temperature, make a contribution to the level of friction. However, most of the friction increase is believed to arise in the quasi-fluid nano particle layer at the pad surface and the real area contribution connected to a quasi-fluid surface. By impeding the creation of secondary plateaus, the reduced area of secondary plateaus leads to a lower average coefficient of friction. One possible way achieve such a plateau destroyer is to roughen the disc surface by e.g. grit blasting, see Fig. 10. This enables grit blasting as a tool for investigating the influence of the coefficient of friction at different locations of the brake disc [9]. 21.

(163) When roughing the brake disc by grit blasting, the formation of secondary plateaus will be impeded in several ways. Initially, sharp asperities on the surface directly ruin any formation. A number of brakings later all protruding roughness has become worn flat. However, the pits are still left. These pits may act as particle traps, thus draining some debris from the interface. More importantly, the fragile plateaus may easily spall of and brake when a pit passes it. The plateau is primarily held in place by the pressure from the disc. When this pressure is lost, the plateau will rise slightly due to elastic spring back. When the trailing edge of the pit hits the thin plateau flake the risk of detachment or fracture is high.. Figure 10. a) The entire friction surface has been grit blasted with 400 µm grits giving the surface an Ra of 9 µm. The disc has not been run after the grit blasting procedure and is full of sharp edges and pits. b) the surface of an untreated disc runin for 420 brakings. The untreated disc has an Ra of 0.5 µm.. Fully grit blasted and untreated discs When braking a grit blasted disc, the coefficient of friction is initially significantly lower than for an untreated disc, see Fig. 11. and Fig. 12. Eventually the grit blasted pits get worn flat or filled with compacted debris and the coefficient of friction slowly returns to the level of an untreated disc. Simultaneously with the reduction of the coefficient of friction the squeal propensity is affected. When grit blasting the whole disc the squeal disappears for a number of brakings but it gradually returns until it has reached the squeal tendency of an untreated disc. No squeal was generated when the coefficient of friction was below 0.4. The frictional level under which no squeal was detected is called the frictional squeal threshold and the phenomena is referred to as the friction threshold effect. By using this effect, experiments studying the squeal propensity of different grit blasted patterns were possible. The squeal index (SI) is used as an indicator of the squeal propensity of a brake disc. This is the number of all squealing points measured divided by the total number of measured points. The untreated disc in Fig. 12 has a SI=0.60 which is typical for an untreated disc under these test conditions. 22.

(164) Figure 11. The friction development of a fully grit blasted disc. The coefficient of friction is slowly increasing and around the 200th braking the first squeal occurs. The brake sequence used is 42 brakings long with brake line pressures from 3 to 24 Bar. Higher brake line pressures give a higher coefficient of friction, which explains some of the short term variations in the graph.. Coefficient of friction. 0,8 Squealing brakings Silent brakings. 0,7 0,6 0,5 0,4 0,3 0,2 0,1 0 0. 50. 100. 150 200 250 Number of brakings. 300. 350. 400. Figure 12. The friction development of an untreated disc. The squealing measurements generally have higher friction than do silent brakings. Squeal does not appear for brakings with coefficient of friction below 0.4. The brake sequence used is 42 brakings long with brake line pressures varying between 3 and 24 Bar. Higher brake line pressures give a higher coefficient of friction, which explains some of the short term variations in the graph.. 23.

(165) Concentric grit blasted patterns (Paper II) Three discs were grit blasted with concentric circles, see Fig. 13.. Figure 13. Concentric patterns grit blasted on three brake discs called outer ring, middle ring and inner ring. Dark areas symbolize grit blasted areas. The actual treated area differs between the discs, from 34% for the outer ring disc to 28% and 21% for the middle ring and the inner ring disc respectively.. The result was that the outer ring disc produced less squeal than an untreated disc, the middle ring disc equally as an untreated disc and finally the inner ring disc, which squealed more than an untreated disc, see Fig. 14. The newly grit blasted areas act like sandpaper and cause, due to the pronounced abrasive roughness around the individual craters, a more rapid wear of the pad. The abrasive contact against the disc will also impede the formation of large and stable contact plateaus [9], which results in a lower coefficient of friction against the grit blasted areas of the disc. It should be noted that there is a fundamental difference in the contact and wear conditions between fully and partly grit blasted discs, which will result in different friction behaviour. In the case of fully blasted discs, the rapid wear is evenly distributed over the pads and does not lead to any redistribution of the pad pressure. If only a concentric ring of the disc is grit blasted, this will initially lead to rapid wear of the corresponding area on the pad. Soon, however, the local removal of material will reduce the local pressure, which accordingly will reduce the local wear rate. After some time the wear rate over the blasted area will match that of the rest of the pad at which point the pressure situation stabilises. The pressure distribution over the pad will then be in direct proportion to the wear resistance of the pad against the corresponding disc surface. The test with the concentrically grit blasted circles showed that depending of the location of the treated area, the squeal propensity could be dampened as well as amplified. Some of the change in squeal propensity could be referred to the friction threshold effect, but the threshold effect alone is insufficient to give the whole explanation to the phenomena. This was, except depending on the roughness it self, suggested to be an effect of the higher stiffness of the disc closer to the hub and thus a higher tendency to vibrations in the periphery. When applying a larger portion of the total friction force and clamping pressure closer to the hub by grit blasting the outer ring pattern, the 24.

(166) squeal propensity was clearly reduced but never silenced. On the other hand, when the grit blasted area was located close to the hub and, hence, a larger portion of the friction force further out at the disc, the squeal got worse. This weaker effect could be called the pressure displacement effect, giving different squeal propensity depending on the localization of the surface treatment. 0,7. Squeal index Friction. Squeal index and coefficient of friction. 0,6. 0,5. 0,4. 0,3. 0,2. 0,1. 0 Fully blasted. Outer. Middle. Inner. Untreated. Figure 14. The average squeal index and coefficient of friction over 420 brakings for the concentrically grit blasted brake discs plus the fully grit blasted disc and the untreated disc.. All these discs have static patterns, which mean that the grit blasted part of the discs constantly affects the same area of the pads. The opposite is dynamic patterns which cause repeatedly varying conditions between treated and untreated counter surface. To further investigate the significance of the pressure distribution on the squeal propensity a number of dynamic patterns were designed. These were produced and tested with varying degree of success. Spirals and other dynamic patterns (Paper III, IV and V) By introducing a pattern like the 8 spiral arms pattern, or any of the other patterns in Fig. 15, every point of the pad surface switch between contact against treated and untreated disc surfaces.. 25.

(167) b). a). d). c). e). Figure 15. Sketches of the textured patterns used. a) 8 spiral arms, b) long spiral, c) spotted, d)2-spoke e) 8-spoke. The discs rotate clockwise.. All the discs showed average coefficient of frictions well above the squeal threshold of µ = 0.4 found for the fully grit blasted disc. Only a small friction dip is seen for the first few brakings of the spotted and the 8 spiral arms discs see Fig. 16. The two previously described silencing effects should not be expected to be active here. The friction is well above the threshold of the friction threshold effect, and we have no pressure displacements from the periphery towards the hub. However, when comparing the squeal index development of the discs a significant difference is obvious, see Fig. 17. Here, an additional a long term effect is acting. In contrast to the friction threshold effect and the pressure displacement effect, that both gave the strongest anti-squeal effect in the beginning, before the pattern was severely worn, the 8 spiral arms pattern has a run-in period of up to 1000 brakings before entering a non-squeal mode. The anti-squeal effect is held for a period of 1700 brakings, and then gradually vanishing. This is a long time effect compared to the fully grit blasted disc that started squealing after less than 200 brakings.. 26.

(168) Coefficient of friction. 1 Untreated disc. 0,8 0,6 0,4 0,2 0 0. 500. 1000 1500 2000 2500 3000 Number of brakings. 3500. 4000. Figure 16. The average coefficient of friction of the dynamically patterned discs 2spoke, spots, long spiral and 8 spiral arms compared to that of an untreated disc. The coefficient of friction is stable for all of the tested patterns and is well above the friction squeal limit of µ=0.4 during the whole test of up to almost 4000 brakings.. 1. 0,8. Squeal index. Untreated. 0,6. 0,4. 0,2. 0. 0. 500 1000 1500 2000 2500 3000 3500 4000 Number of brakings. Figure 17. The squeal index of the four dynamic pattern discs in the long time test. The disc with two spokes pattern squeals approximately as much as an untreated disc. The spotted disc and the long spiral disc show the same basic squeal behaviour while the 8 spiral arms disc is almost totally silent for a long time. The low squeal period remains for 1700 brakings. After 4000 brakings all discs except the 2-spoke disc have reached the same squeal index of 0.55.. Measurements of the pressure distribution were made with pressure sensitive foils, which is a unique way to get a picture of the pressure distribution over 27.

(169) a whole pad. The same method was also used to follow the evolution of the pressure distribution of the pads. The technique revealed that the grit blasted spiral arms became increasingly elevated above the rest of the disc, see Fig. 18. Surface profile measurements made confirmed the pattern elevation.. a). b). c). Figure 18. Pressure sensitive foils show the evolution of the pressure situation of the 8 spiral arms disc. High intensity of colour indicates a high pressure region in the pad-disc contact. High pressure regions caused by the elevated spiral arms are marked by arrows in b). In a) the elevation has not yet formed and in c) it has become worn off. a) After 210 brakings, b) after 1500 brakings and c) after 3780 brakings. In b) two stripes marked out with arrows show the elevated spiral arms. These stripes are missing in a) and c).. There is a clear correlation between the pattern height and the reduction in squeal propensity, see Fig. 19. When the pattern gets high enough the mechanism generating the squeal is inhibited by the forced vibrations caused by the pattern elevation. In this particular case the squeal is damped considerably for pattern heights of about 25 µm. A squealing brake disc is typically vibrating at frequencies of around a few thousand kilohertz, in amplitudes of perhaps up to a few micrometers [20].. 28.

(170) 0,8. Pattern Height. 40 35. Squeal Index Untreated Disc. 15. 0,2. 10. 0,1. 5. 0. 0. 500. 0. Number of Brakings. 4000. 0,3. 3500. 20. 3000. 0,4. 2500. 25. 2000. 0,5. 1500. 30. 1000. 0,6. Pattern Height [µm]. Squeal Index. 0,7. Squeal Index. Figure 19. The squeal index of a spiral patterned disc compared to the average spiral pattern height. At first the squeal generation is rather frequent but after approximately 1000 brakings the squeal index is below 0,1. This low squealing mode is maintained for about 2000 brakings. The drop in squeal generation is synchronized with the build up and breakdown of the spiral arms at the treated surface.. It is interesting that the 8 spiral arms pattern is the only pattern with a strong influence on the squeal even though all patterns become elevated and thus causes travelling pressure concentrations in the pad-disc contact. The 8 spiral arms pattern causes a pressure wave to move forward-outwards over the pad while the long spiral pattern gives a pressure wave with a more outwards sweeping motion and the spoke patterns only a forward motion of the pressure wave. Even though the exact mechanism behind the squeal reduction still remains to identify and understand, a good explanation to the mechanism behind the pattern elevation can be given. Particle embedding The gradual elevation of the grit blasted dynamic patterns can be explained by the effect the grit blasting has on the wear resistance of the disc material. SEM images of grit blasted discs show traces of silicon carbide (SiC), the erodent used to roughen the surface, see Figs. 20 and 21. Some of these grit fragments are embedded in the disc surface and act as wear resistant inclusions, see Fig. 22. After some brakings the thus wear protected pattern starts to rise above the surface and this elevation continues until the wear resistant particles are all worn away. At this point the elevated area will start to wear more rapid than the surroundings, due to the higher pressure. The wear pro29.

(171) tection from the SiC fragments starts vanishing after 2000 brakings. If the wear resistant particles could be injected deeper into the disc, the squeal could be suppressed for a considerably longer time. One way to achieve deep particle injection is to locally melt the disc material in the pattern and inject a hard phase into the melted material. This is possible using laser particle injection.. Figure 20. Fragments of SiC particles injected into the disc surface by grit blasting.. Figure 21. Combined SEM-EDS image of a treated disc surface after 1050 brakings. Dark areas are SiC fragments embedded in the cast grey iron.. 30.

(172) Figure 22. The grit blasted disc surface with an embedded SiC particle. The fragment has become firmly integrated and worn flat by the contact against the pad. The pit was produced by using FIB technique.. Laser particle injected disc In the experiments where laser particle injecting was used as an alternative to grit blasting, a laser draws the pattern onto the disc melting the grey iron. Simultaneously tantalum carbides are injected into the melt. As soon as the laser spot has passed, the surrounding iron act as a heat sink quenching the melted iron producing a hard martensite with very hard tantalum carbides well embedded. The laser particle injected disc already from the beginning showed an elevated pattern because of the volume added by the injected particles. The pad wear was initially quite severe but eventually it returned towards the level of untreated discs. The average coefficient of friction was below the friction threshold, but single measurements were above the limit. No squeal was though detected during the 3400 brakings the disc was run, see Fig. 23. In combination with the very efficient elevated 8 spiral arms pattern the low coefficient of friction give a totally squeal free brake disc.. 31.

(173) Coefficient of friction. 1 Coefficient of friction Squeal index. 0,8 0,6 0,4 0,2 0 0. 500. 1000. 1500 2000 2500 Number of brakings. 3000. 3500. 4000. Figure 23. The coefficient of friction and the squeal index for the laser particle injected disc. 10% of the measured points are displayed in the graph. The squeal index equals zero during the whole test and although the average coefficient of friction is below the friction threshold of 0.4, about 3.5 % of the brakings have a coefficient of friction above the threshold.. 32.

(174) Gripping surface (Paper VI) As an expansion of the work by Pettersson and Jacobson [21], the highfriction properties of textured diamond surfaces designed for gripping has been investigated. The diamond surfaces are equipped with sharp, extremely well defined pyramids. The well defined very sharp pyramids offer extremely high static coefficient of friction, up to 1.6 for certain geometries. These surfaces have been mated against different metal surfaces and the static coefficient of friction has been measured under varying contact conditions. The diamond surface was produced using micro mechanical techniques based on photolithography and anisotropic etching of silicon followed by chemical vapour deposition. By etching pyramid shaped pits in the silicon, depositing diamond, electroplating with nickel and finally etching away the remaining silicon, a surface equipped with thousands of protruding diamond pyramids is created [22, 23], see Fig. 24.. Figure 24. Diamond surface with a pyramid texture. Each pyramid has a base line of 180 µm and the height is 85 µm. The area fraction of pyramids is 22%. SEM.. The pyramid geometry allows friction testing in two basic directions; either with a corner in the sliding direction or with a face in the sliding direction. Theoretically the corner first orientation would give a higher coefficient of friction than the face first orientation since the ploughing area is larger for the same load carrying area. Transition from pyramid only contact to mixed contact At low loads only the outermost pyramids tips indents the surface and all the load is carried directly by the pyramids. At higher loads, the pyramids indent deeper and some contact may appear outside the pyramids. In a simplified, idealised situation, where the pyramid surface contacts a flat, perfectly aligned surface that does not form pile-ups around the pyramid indents, no contact outside the pyramids would occur, until the pyramids were fully indented into the flat surface. However, in any practical situation a number of other deviations should also be expected to reduce this threshold load, as indicated in Fig. 25. 33.

(175) Figure 25. Alternative contact situations between the microtextured diamond surface and metallic countersurfaces. a) Idealised contact situation, without pile ups, at low load where the entire load is carried by the pyramids. b) Idealised contact situation, without pile ups, at high load where load is also distributed outside the pyramids. c) Situation with pile up, where the ridges carry part of the load. d) Situation with imperfect alignment. e) Situation with rough metallic surface.. As soon as the flat diamond carries load, the friction force FT will not continue to grow proportionally to the load. It now becomes the weighted mean value between the pyramid ploughing value µp and the flat against flat value µa.. FT. P p FN , p P a (FN – FN , p ). (Eqn. 1). Where FN is the total load and FN,p is the load carried directly by the pyramids. This equation is plotted in Fig. 26 for a number of different area fractions of pyramids. The larger fraction that is covered by pyramids, the higher the critical load for contact against the flat surface.. 34.

(176) Figure 26. Friction situation in contacts varying from pyramid contact only to a mixed pyramid and flat contact. Curves corresponding the idealised situation described by Eqn. 1. (Assumes smooth, well aligned surfaces, without pile-ups, etc.) a) Friction force as a function of the load. b) Coefficient of friction as a function of area fraction in contact.. These details regarding the practical limitations of the use of designed high-friction surfaces where further investigated in Paper VI. The influence of pile-up was modelled, using a simplified geometry of the ridges formed. According to this model the pile-up would reduce the maximum load for pure pyramid contact to about 50% of the load predicted without ridge formation. Further, as soon as a tangential load is applied, the situation gets worse. The surfaces might tilt slightly, and more material pile up in front of each pyramid, initiating the formation of a prow or a chip, see Fig. 27.. Figure 27. Scratches formed in the silver surface at different loads. At low loads the ridges and prows formed are too small to contact the flat surface outside the pyramids. At high load the ridges rise high enough to contact the flat surface and thus relieve some load. a) Nominal pressure 0.7 MPa. b) Nominal pressure 24 Mpa.. The load limit was investigated by using high loads and sparse diamond patterns, to cause really deep penetration. The predicted load dependence 35.

(177) was found to hold qualitatively. However, for the diamond structure with pyramids covering 22% of the surface, the coefficient of friction was found to start falling at nominal pressures corresponding to around 5% contact rather than around 11% as modelled, see Fig. 28. This was also the case for the other tested geometries, which likewise started falling well before the point predicted by the simplified model. This discrepancy was suggested to be due both to increased pile up during the initial movement, non-perfect surface alignment and non-flat surfaces, factors that are not included in the model.. Coefficient of friction. 1,4. Flat Dense, corner first Dense, face first Sparse, corner first. 1,2 1 0,8 0,6 0,4 0,2 0 0. 0,05. 0,1. 0,15. 0,2. 0,25. Real contact area fraction. 0,3. 0,35. Figure 28. Frictional behaviour of four diamond surfaces mated against silver. The load has been varied to vary the fraction of real counter area. Flat: diamond surface without pyramids. Sparse: Diamond surface with 2 % area covered by pyramids. Dense: Diamond surface with 22 % area covered by pyramids. All tests have been run under well lubricated conditions. Compare with Fig. 26 b).. In a related investigation the effect of repeated use was investigated. The technically interesting question is whether the surfaces can be repeatedly loaded, without affecting the level of friction. Specially, the surfaces are quite severely deformed each time a relative movement is induced. This could be suspected to ruin the possibilities to disassemble and reuse a coupling surface, as soon as it has once lost its grip. It was found, however, that the friction coefficient is remarkably insensitive to this effect, see Fig. 29. The only exception was the highly loaded silver surface that exhibited a gradually decreasing coefficient of friction, with increasing number of slip events. The surfaces may be substantially altered without seriously affecting the friction as obvious from the examples in Fig. 30.. 36.

(178) 1,6. Coefficient of friction. 1,4. 1,2. 1. 0,8. 0,6. steel 20 MPa steel 3.4 MPa silver 3.4 MPa silver 20 MPa. 0,4. 0,2. 0 0. 2. 4. 6. 8. 10. 12. Number of slip events. Figure 29. Influence on the static coefficient of friction from repeated overloading, resulting in relative motion. The surfaces were separated, slightly repositioned and mounted, between each slip event.. 37.

(179) a). b). c) Figure 30. Surfaces after slip events resulting in loose of grip and relative movement. The surfaces were slightly repositioned between each event. Only the surface in b) resulted in a somewhat reduced friction (about 20% reduction). a) Silver surface after 5 slips. Nominal pressure 3.4 Mpa. b) Silver surface after 5 slips. Nominal pressure 20 Mpa. c) Steel surface after 10 slips. Nominal pressure 20 MPa.. 38.

(180) Effect of surface preparation In a further attempt to investigate the sensitivity of the surface characteristics, a series of different roughnesses were prepared by grinding steel samples with papers of different meshes. The series of paper meshes 120, 220, 320, 500, 1200 and 4000 resulted in Ra values of 1.84, 0.75, 0.67, 0.55, 0.23 and 0.13 µm respectively. Despite the wide range in Ra values the coefficient of friction was virtually unaffected, see Fig. 31.. Coefficient of friction. 1,6 1,4 1,2 1 0,8 0,6 0,4. Along grinding direction Across grinding direction. 0,2 0 0. 0,5. 1. R [µm]. 1,5. 2. a. Figure 31. Coefficient of friction of the dense diamond structure against steel of varying surface roughness. A nominal pressure of 3.4 MPa is used.. The facts that these surfaces were remarkably stable at high friction levels, and that a maximum load limit was found above which the coefficient of friction falls are among a number of other factors found important for the successful design of high-friction joints.. 39.

(181) Summary in Swedish. Bland det första en materialforskare med intresse för friktion och nötning – det ämne vi kallar tribologi – måste få klart för sig är att en materialyta sällan besitter samma egenskaper som bulkmaterialet. Man kan exempelvis lätt tänka sig att ett stålföremål har en yta av stål, men i verkligheten täcks stålet av andra ämnen och föreningar som bland annat ändrar ytans kemiska och tribologiska egenskaper. Först och främst finns där alltid en metalloxid, och utanpå den hittar man bland annat adsorberad vattenånga, organiskt material i form av t.ex. bakterier eller fett från fingrar som tagit på ytan. Ofta har dessa främmande substanser en smörjande och nötningsskyddande funktion, vilket oftast är positivt. I andra fall kan det vara oönskat eftersom det i många applikationer, såsom i fordonsbromsar och greppytor, är vitalt med en hög och stabil friktion för bibehållen funktion. Friktion definieras som den reaktionskraft som hindrar relativrörelse mellan två ytor i kontakt. Friktion är i sig inte en materialegenskap. Den påverkas av materialegenskaper som hårdhet, smälttemperatur, o.s.v., men det går alltså aldrig att bestämma friktionskoefficienten för ett visst material eftersom friktionen är resultatet av en interaktion mellan ytor vars verkliga sammansättning vi inte känner och som ändras med tid och med pågående nötning. Dessutom inverkar ytornas geometri kraftigt på friktionen. En slät yta ger oftast en lägre friktion än en ojämn yta till följd av den geometriska låsning som yttopparna ger. Friktionskoefficienten, µ, beräknas utifrån den pålagda kraften och friktionskraften enligt sambandet:. µ. Friktionskraft Normalkraft. När två ytor gnids mot varandra kan ett material med andra egenskaper och annan sammansättning än de två ursprungliga materialen bildas i kontakten. Detta nya material kallas tredjekroppslager och består ofta av oxiderade nötningsfragment från originalmaterialen.. 40.

(182) Bromsar Friktionsytorna i en bilbroms består idag dels av en bromsskiva gjord av perlitiskt gråjärn, och dels av bromsbelägg innehållande en blandning av upp till 30 olika ingredienser. En grovindelning av ingredienserna brukar ske i följande grupper: x Strukturmaterial vilka har till uppgift att ge mekanisk styrka åt belägget. Hit hör t.ex. stål- och glasfibrer. x Utfyllnadsmaterial vilka ska separera fibrerna och vara temperaturtåliga. Ofta används lermineral som bariumsulfat och vermiculit. x Bindemedel som ska fungera som lim mellan de andra ingredienserna. I de flesta fall är detta någon form av fenolharts. x Friktionsadditiv har i uppgift att smörja kontakten mot skivan för att få en stabil friktion. Även keramiska partiklar tillsätts för att slipa bromsskivan ren från rost. Plattåmodellen Kontaktsituationen mellan skivan och beläggen är väldigt komplex, har beskrivits i Plattåmodellen, som vi arbetat med att förfina. Den verkliga kontakten mellan två fasta kroppar sker i små mikroskopiska kontaktpunkter, och detta gäller även för kontakten mellan skiva och belägg. Den speciella sammansättningen hos beläggen, med sina extrema skillnader mellan de ingående materialen, gör situationen speciell. De minst nötningståliga faserna nöts snabbt bort, och lasten kommer att bäras på de nu uppstickande nötningsbeständiga faserna, såsom stålfibrerna. Dessa nöts av så att det uppstår ett antal glidplatåer, spridda över belägget. Vi kallar dessa för primära kontaktplatåer. Utanför dessa platåer finns ett litet utrymme mellan skivan och belägget i vilket små nötningsfragment kan ta sig fram, se Fig. 1. Dessa nötningsfragment kan dock stocka sig mot en primärplattå, och bilda en hoppackad platta av nötningspartiklar, som även den kan bära en del av lasten. Vi har då bildat en så kallad sekundärplattå.. Figur 1. En schematisk bild av kontakten mellan skiva och belägg. Gråa nötningsfragment tar sig fram i labyrinten mellan skiva och belägg men kan fastna mot vita primärplattåer och bilda sekundärplattåer.. 41.

(183) Genom att med en jonstråle skära ut en liten bit av en sekundärplattå och studera denna i ett transmissionselektronmikroskop (TEM) har vi kunnat avslöja nya detaljer om plattåns uppbyggnad, se Fig. 2.. Figur 2. Vänstra bilden visar en sekundärplattå av hoppackade järnoxidpartiklar. De svarta fragmenten 1 till 2 µm under ytan består av järn. (Den lilla) fyrkanten visar det förstorade området till höger. Skalstrecket i den högra bilden är 50 nanometer långt vilket motsvarar ca 250 atomer.. I delförstoringen syns två stora gråa fält avdelade en mörk gräns emellan. Det övre gråa fältet är ett nanokristallint deformationsskikt. Detta byggs upp av nanometerstora partiklar som mixas runt av kraften och skjuvningen från skivan. Nedanför den svarta gränsen finns material som inte deltar i omrörningen i samma omfattning som materialet ovanför. Tillväxten av nya sekundärplattåer under varje ny bromsning bidrar till att friktionskoefficient mellan skiva och belägg ökar under varje längre inbromsning, se Fig. 3.. Friktionskoefficient. 0,7. Sista mätpunkten. En bromsning. 0,65 0,6 0,55 0,5 0,45 0,4 200. Första mätpunkten. 201. 202 203 Antal bromsningar. 204. 205. Figur 3. Under den inringade bromsningen ökar friktionskoefficienten från 0.47 i den första mätpunkten till 0.62 i den sista mätpunkten, till följd av skapande och tillväxt av sekundärplattåer på belägget.. 42.

(184) Bromsskrik En oönskad egenskap hos många bromsar är de skrikande ljud som kan uppstå vid en inbromsning. Vi har kunnat eliminera dessa oljud genom att blästra bromsskivans slityta. Ljudet kom dock tillbaka efter några hundra bromsningar. Anledningen till detta är att det finns en lägsta nivå för friktionskoefficienten mellan skiva och belägg under vilken skrik inte genereras. När skivan är nyblästrad förhindrar dess ojämna yta att sekundärplattåerna växer till, vilket gör att friktionen sänks. Efter ett tag har det blästrade mönstret nötts så att friktionen åter kommer över denna lägsta friktionsnivå och skriken återkommer. Vi kallar detta friktionströskeleffekten och vi har använt den som verktyg för att utröna om olika friktion på olika delar av skivan ger inverkan på skrikförmågan. Tre bromsskivor blästrades med cirkelformade mönster, se Fig. 4. Figur 4. Cirkelformade mönster blästrade på bromsskivor. De mörka områdena symboliserar blästrad yta.. Det visade sig att det spelade roll var denna cirkel blästrats. Om cirkeln blästrats i mitten skrek det lika mycket som en oblästrad skiva, men om den yttre cirkeln blästrats skrek det lite mindre och för den inre cirkeln lite mer. Detta beror på att skivan är styvare närmare centrum. Där skivan blästras kommer beläggen till en början att nötas snabbare på grund av den blästrade ytans slipande effekt. Trycket från belägget mot skivan omfördelas då till de oslipade delarna, och om trycket appliceras närmare centrum skriker det mindre medan det skriker mer om trycket läggs långt ut på skivan. Vi kallar detta för tryckförskjutningseffekten. Några andra mönster som saknar tryckförskjutningseffekt testades i en annan undersökning, se Fig. 5. a). b). c). d). Figur 5. a) spiral, b) lång spiral, c) prickar, d) streck. 43.

(185) Dessa visade sig ha väldigt olika inverkan på skriket, och till skillnad från de två tidigare nämnda effekterna märktes denna antiskrikeffekt inte till en början, utan det kunde ta flera hundra bromsningar innan skrikandet avtog, se Fig. 6. 1. Streck Lång spiral Spiral Prickar. 0,8. Skrikindex. Obehandlad skiva. 0,6. 0,4. 0,2. 0. 0. 500 1000 1500 2000 2500 3000 3500 4000 Antal bromsningar. Figur 6. Y-axeln är märkt ”Skrikindex” vilket är ett jämförelsevärde där 1 betyder att det skriker hela tiden och 0 att det är helt tyst. För den spiralblästrade skivan minskar skriket ner till 0 för att senare öka till över 0.5. De andra mönstren har inte samma effekt på skriket.. Det visade sig att fragment från blästermedlet fastnade på skivans yta och gav mönstret en högre nötningsbeständighet än resten av skivan. Detta resulterade i att mönstret blev upphöjda åsar vilka ledde till en ny kontaktsituation, med ständigt växlande tryckfördelning mellan belägget och skivan. Tryckväxlingarna tvingar på skivan en rörelse som ersätter den svängning som ger skrik. När mönstret nötts bort kommer skriket tillbaks igen.. Greppytor I vissa applikationer är det viktigt att komponenter inte glider i förhållande till varandra, men att de ändå är demonterbara. Då används ofta t.ex. skruvförband där bultar används för att hålla samman ytorna. Genom att ha en hög friktionskoefficient mellan ytorna kan bultkraften minskas utan att förlora i skjuvmotstånd. Färre bultar kan då användas vilket sparar vikt och ger lägre kostnad för konstruktionen. I denna undersökning har texturerade diamantytor använts för att studera friktionsmekanismer vid övergången mellan ren plogande friktion och blandad adhesiv och plogande friktion hos metall, se Fig. 7. 44.

(186) Figur 7.Diamantyta med pyramidtextur. En pyramid har en sida på 180 µm och är 85 µm hög. Ytandelen pyramider är 22%.. Pyramidformen ger tack vare det linjära sambandet mellan plogande och lastbärande area en konstant friktionskoefficient. När hela pyramiden tryckts in i otytan och den plana ytan mellan pyramiderna börjar bära last ändras detta förhållande. En ideal kontakt mellan en diamantpyramidyta ser ut som i Fig. 8 a) och b). Pyramiderna tränger in i motmaterialet tills samtliga pyramider är helt intryckta. Därefter börjar den plana diamantytan bära last. I verkligheten bildas intrycksvallar i metallen runt varje pyramid. Dessa vallar växer ju djupare pyramiden trycks in, men till sist når plogvallen till den plana diamantytan och börjar bära last, Fig. 8 c). Den plogande arean är nu mindre än i det tänkta idealfallet, vilket medför en sänkt friktionskoefficient. Andra effekter som kan sänka friktionskoefficienten är om motytan inte är planparallell med pyramidytan eller om motytan har en avsevärd topografi, Fig. 8 d) och e).. Figur 8. Olika kontaktsituationer mellan en texturerad diamantyta och en metallmotyta. a) Idealiserad kontakt situation, utan vallbildning, vid låg last när hela lasten bärs av pyramiderna. b) Idealiserad kontakt situation, utan vallbildning, vid hög last när last bärs av både pyramider och plan yta. c) Situation med vallbildning, där vallarna bär en del av lasten. d) Situation med sned motyta. e) Situation med grov motyta.. 45.

(187) Idealt skulle friktionskoefficienten vara konstant medan pyramiderna bär lasten, dvs fram till en areaandel av 22 % i Fig. 9. Sedan minskar friktionskoefficienten ju mer last som bärs av den plana ytan. I verkligheten ger avvikelserna från den ideala situationen att friktionskoefficienten ger vika vid betydligt lägre areaandel än 22 %. Redan vid en areaanel på 5 % börjar friktionskoefficienten minska. Detta gäller oavsett om diamantpyramiderna orienterats med ett hörn eller en sida i dragkraftens riktning. Plan diamant 22 % areaandel, hörn framåt 22 % areaandel, sida framåt 2 % areaandel, hörn framåt 1,4. Idealfall, 22 % areaandel, hörn framåt. Friktionskoefficient. 1,2 1 0,8 0,6 0,4 0,2 0 0. 0,05. 0,1. 0,15. 0,2. Verklig ytandel. 0,25. 0,3. 0,35. Figur 9. Belastade areaanelens inverkan på friktionskoefficienten. En yta som idealt sett ska ha konstant friktionskoefficient fram till en belastad areaandel på 22 % ger i själva verket konstant friktionskoefficient endast till en belastad areaandel på 5 %. Detta till följd av vallbildning, snedbelastning och ytojämnhet på motytan.. 46.

(188) Acknowledgements. This work has been carried out at the Tribomaterials group, The Ångström Laboratory, Uppsala University. This work has been financed by the Volvo AB in cooperation with the Swedish taxpayers through the Swedish Board for Industrial and Technical Development. I would like to acknowledge Axel Hultgrens Fond, Liljewalchs resestipendium and Duroc Energy AB. I would like to thank all my past and present co-workers at the Uppsala University. Especially:. Staffan Jacobson for his astonishing ability to turn science into pure poetry, and for nice company during our visit in Japan. Sture Hogmark for giving me the opportunity to be a part of the Tribomaterials group for five short years, and that he never run short of good stories. Carin Palm, Janne Gustafsson, Rein Kalm och Caroline Olofsson for being the perfect support team without no one at the division could survive. Urban Wiklund for our excellent cooperation in teaching students the secret of materials. Joakim Andersson, Patrik Hollman, Ulrika Pettersson, Ernesto Coronel, Åsa Kassman, Filip Bergman, Mikael Åstrand, Claes Kuylenstierna and Lars Claesson for all your help and expertise. Claes Aldman, Verkstads-Åke, Larsa and Degen for the small talk. Daniel Persson, Marcus Lehto, Nils Stavlid, Magnus Hanson, Mattias Carlsson, Julia Gerth, the Mikromeckarna, Sophie Ohlsson, and all the project and diploma workers for all fun and stupid conversations. A special thanks to those of you that took time to proof read! My buddies Ulrik Beste and Martin Nilsson for support on and off the track. My brother David, Helena, Ludwig, my sister Merete, Micke, Elias, Josefin and my god daughter Line for always being there for me! My parents Ethel and Fredrik. For everything! Words are not enough! My beloved wife Christine for all your love and support! I love you! My son Filip for being the centre of my universe and for dragging me out of bed 05.40 in the morning! I love you! Uppsala 2005-05-18 Lars Hammerström 47.

(189) References. 1. Blau, P.J., The significance and use of the friction coefficient, Tribology international 34, (2001), 585-591. 2. Godet, M., The third-body approach: a mechanical view of wear, Wear, v 100, n 1-3, (1984), 437-452. 3. Hogmark, S., Jacobson, S., Wänstrand, O., A new universal test for tribological evaluation, Proceedings of the 21st IRG-OECD Meeting, Amsterdam, Netherlands, 1999. 4. Hogmark, S., Jacobson, S., Wänstrand, O., The Uppsala load scanner-an update, Proceedings of the 22nd IRG-OECD Meeting, Cambridge, United Kingdom, 2000. 5. Bergman, F., Eriksson, M., Jacobson, S., A Software Based Measurement System for Test and Analysis of Automotive Brake Squeal, TriboTest journal, v 5, n 3, (1999), 265–275. 6. Jud, P.P., Nellen, P.M., Sennhauser, U., Micromachining by Focused Ion Beam (FIB) for materials characterization, Advanced Engineering Materials, v 7, n 5, (2005), p 384-388. 7. Jacobson, S., Hogmark, S., Tribologi. Karlebo-serien, 1996, Liber Utbildning AB, ISBN 91-634-1532-1. 8. Kiple, R.C., et al., Thermal cracking in disc brakes, Engineering Failure Analysis, v 9, n 1, (2002), p 63-76. 9. Eriksson, M., Bergman, F., Jacobson, S., On the nature of tribological contact in automotive brakes, Wear, v 252, n 1-2, (2002), p 26-36. 10. Eriksson, M., Jacobson, S., Friction behaviour and squeal generation of disc brakes at low speeds. Proc. Instn. Mech. Engrs. Part D; J. of Automobile Engineering, v 215, n 12, (2001), 1245-1256. 11. Rhee, S.K., Tsang, P.H.S., Wang, Y.S., Friction-induced noise and vibration of disc brakes, Wear, v 133, n 1, (1989), p 39-45. 12. Kinkaid, N.M., O'Reilly, O.M., Papadopoulos, P., Automotive disc brake squeal, Journal of Sound and Vibration, v 267, n 1, (2003), p 105-166. 13. Eriksson, M., Jacobson, S., Tribological surfaces of organic brake pads, Tribology International, v 33, n 12, (2000), p 817-827. 14. Eriksson, M., Lord, J., Jacobson, S., Wear and contact conditions of brake pads: Dynamical in situ studies of pads on glass, Wear, v 249, n 34, (2001), p 272-278. 15. Österle, W., Urban, I., Friction layers and friction films on PMC brake pads, Wear, v 257, n 1-2, (2004), p 215-226. 48.

No results found