EP3001256B2 - Anchor escapement - Google Patents

Description

The invention relates to an anchor escapement for a mechanical timepiece comprising a pivotable anchor with anchor pallets and an escape wheel which can be subjected to a torque and which has teeth directed approximately radially outwards over its outer circumference. The front flanks of the pallets are designed in such a way that they always engage with the same contact surface every time they come into contact with the front flank of the teeth of the wheel.

The escapement of a watch is the assembly in wheeled watches that creates the connection between the gear train and the gear regulator. It usually consists of the escape wheel and the anchor piece. The gear regulator causes the gear train to periodically stop (inhibit) via the brake piece that engages in the escape wheel and thus the regular movement of the watch.

Various inhibitions have been proposed in the prior art. In the meantime, practically all mechanical watches are equipped with the same type, namely the so-called “Swiss anchor escapement”. In a "Swiss anchor escapement", the two arms of the anchor each contain an anchor stone (pallet), which is usually made of ruby, sapphire or garnet. These anchor stones are either inserted into the two arms of the anchor or, in the case of micro-engineered anchors, are made from silicon or diamond in one piece together with the anchor. The anchor stones alternately grip each tooth of the escape wheel and hold it firmly.

The escape wheel and anchor are then at rest and are immediately accelerated again when the balance wheel crosses zero. Every time the balance wheel passes the zero position in one direction or the other, it engages the anchor fork via the so-called lever stone. As a result, the anchor releases one tooth of the escape wheel via its respective pallet, which then advances briefly and returns a tiny fraction of energy via the anchor to the lever stone and thus the balance wheel. When power is transmitted between the teeth of the escape wheel and the pallets of the anchor, these two parts move against each other under pressure. At the beginning of the movement, a pallet rests on a surface of an armature gear, the so-called resting surface. When the pallet moves against the escape wheel, a frictional force occurs. However, the friction between the pallets and the escape wheel can affect the accuracy and lifespan of a watch movement.

A typical escape wheel has around 20 teeth that are equidistantly distributed around the circumference. The anchor typically 2 pallets. With each revolution of the escape wheel, each piston tooth is brought into contact with each pallet. This means that each pallet of the armature experiences n times the wear (with n = number of teeth on the wheel) compared to a piston tooth on the escape wheel. In particular, the friction between the teeth of the escape wheel and the pallets of the armature leads to material removal, i.e. to wear on the contact surfaces of the pallets and escape wheel, which can reduce the accuracy and the parts in question have to be replaced from time to time. This is also the reason why the pallets are made of a harder material, typically ruby.

The escapement converts around 60% of the movement's energy. Friction losses as well as the constant acceleration and braking of the wheel and anchor are primarily responsible for the low efficiency. In the past, attempts were made to minimize the moment of inertia of the wheel. This was achieved, on the one hand, by as high a degree of skeletonization as possible (removing all unnecessary volume components) and, on the other hand, by making the wheel as thin as possible (approx. 100µm). As a consequence, the anchor pallets must have a certain minimum thickness (e.g. 200µm) to ensure that the wheel engages securely in the pallets.

There has therefore been no lack of attempts to minimize this problem. This is how it is done in the FR 1485813 proposed to bevel the teeth of escape wheels to create a contact surface between the teeth and the anchor pallets with a smaller width. Furthermore, in the EP 1 622 826 B1 suggested using other materials, such as silicon or diamond, which are said to have advantages over lubricated escapements due to higher hardness and lower coefficients of friction. To minimize wear and friction, oils are mandatory for conventional escapements with steel escape wheels and ruby pallets.

CH 612 308 A3 relates to a Swiss anchor escapement, which can be manufactured in a simplified manner, including a one-piece, thin, flat lever of equal thickness over the entire body, so that it can be obtained by a punching process and a bending process for the scriber. These two operations can be carried out simultaneously using the same tool.

However, the new materials, silicon and diamond, are harder than steel. If these materials are used as inhibiting components without lubrication, this also results in technical problems. It turns out that the components have a so-called running-in phase. The frictional contact obviously flattens micro-roughness on the surface. This reduces the coefficient of sliding friction and as a result the amplitude of the balance wheel increases. Usually, before the increase in the amplitude, there is first a decrease in the amplitude, which is caused by the accumulation of foreign bodies (e.g. abrasion, dust, organic particles, etc.), which is similar to the principle of a snow plow due to the existing Surface roughness can be incorporated into the surface and increase the coefficient of friction. In order to avoid this effect, the prior art uses EP 2 107 434 proposed to make the contact surfaces of the teeth of the gear as small as possible. The same applies to abrasion, which can also be incorporated into the surface. In the worst case, this effect can even cause the clock to stop. In order to avoid this effect, such components in the prior art are usually subjected to splash lubrication based on phenomenological observations. The remaining thin oil film on the surface evaporates over time and enables an almost constant balance amplitude during the running-in phase. For this reason, components made of silicon, for example, are immersed in a gasoline/oil mixture before being installed in the clockwork.

Studies carried out show that the phenomenologically observed running-in phase is caused by the leveling of a microscopic roughness. Since the new materials, especially diamond, have a significantly greater hardness than steel, a longer running-in phase can be observed as the micro-roughness remaining after processing is leveled out more slowly. In the case of diamond, the wear is so low that even with very smooth surfaces (e.g. 20nm Rms, square roughness, root-mean-squared roughness) a running-in phase of more than 100 days can still be observed. In order to minimize this, a less hard layer can, for example, be additionally applied to the diamond surface (see for example EP 2 236 455 ), which ensures that the friction surfaces in contact become smooth more quickly. Despite using such layers, the running-in period is still around 40 days. The disadvantage here is that the friction properties change over a longer period of time and therefore it is not possible to regulate the watch reliably and quickly (set the correct accuracy).

It is also typical for the use of new materials and microtechnical manufacturing processes that the functional surfaces of the components (the side flanks) can never be aligned completely vertically - in relation to the top and bottom of the component. This is due to the manufacturing technology adapted from the semiconductor industry, in which the components are etched out of a plate (so-called wafer) using photolithography and dry etching processes. As a result, the side flanks of the armature (particularly the pallets) and escape wheel (particularly the contact surfaces of the piston teeth) can be arranged either plane-parallel or opposite. In the second case, the typical structure of the escapement with a thin wheel and a thick anchor pallet means that the wheel always has a defined contact surface, as is the case in the EP 2 107 434 B1 is described. Here, a defined contact surface is understood to mean a contact surface that is in contact every time the components come into contact. Due to the non-vertical arrangement in the case of diamond, this is very small and has the shape of a 2-4µm wide rectangle. This area is also called the geometric contact area A ₀ . This “contact line” extends roughly into the middle of the anchor pallet. Depending on the position of the clockwork, due to the bearing play of the individual components, the relative position of the armature and wheel changes, so that the contact surface of the armature also shifts relative to the wheel. Therefore, a 30-50µm wide wear zone on the anchor is usually visible in the scanning electron microscope. The contact zone of the wheel, on the other hand, is "defined" by definition due to the non-vertical arrangement, regardless of the relative position of the wheel and armature, and is always in contact with one another, regardless of the relative position of the clockwork components. With all classic escapements, the anchor pallet always dominated the running-in phase, which is why it was made from the harder material ruby.

The investigations also showed that the running-in phase is dominated by the contact surfaces and their properties. If you look at the defined contact surface of a piston tooth, which is in the form of a rectangle, you can see that only a fraction of the geometric contact surface A ₀ is actually in frictional contact with the counterpart. This is due to the nanoscale surface roughness present microscopically. If one examines the so-called airfoil proportion T of the defined contact area, ie the ratio of the area A _R actually in contact with respect to the geometric contact area A ₀ , values of T=10%-30% can be obtained at the beginning of the run-in phase and of more when the run-in phase is completed than 50%. Since no further change (wear, removal) of the contact surfaces could be observed after the running-in phase was completed, there is much to suggest that from this point on a tribofilm can form on the diamond surface, to which the reduction in the coefficient of sliding friction and thus the increase in amplitude of the watch can be attributed is.

In order to achieve the shortest possible run-in phase, a wing portion of the defined contact area of more than 50% must be created as quickly as possible. On the one hand, this can be achieved via the manufacturing process by reducing the surface roughness, or by increasing the wear on the defined contact surface.

The object of the present invention is therefore to propose an anchor escapement which is significantly improved compared to the prior art with regard to the friction and wear between the teeth of the gear and the pallets of the anchor, so that the accuracy and the service life of a clock gear train can be improved and at the same time long run-in phases in lubrication-free running can be dispensed with.

The task is marked by the Features of patent claim 1 solved. The subclaims show advantageous further developments.

According to the invention, the defined geometric contact surfaces A ₀ are now exchanged by a changed geometric arrangement, with the aim of fixing the contact surface of the anchor pallets (thus defined contact surface) and keeping that of the wheel variable. The reason for this arrangement is to accelerate the wear of the anchor, contrary to all previous teachings, in order to reduce the duration of the running-in phase. If the anchor has the defined geometric contact area A ₀ , this area must inevitably experience n-fold wear (with n = number of piston teeth of the wheel). Surprisingly, exactly this effect occurs and the running-in phase in the case of diamond almost completely disappears.

The industrial implementation of such an inverted arrangement must also enable safe engagement of the wheel and anchor. As a result, the wheel, or at least the thickness of the piston tooth, must be made significantly thicker than that of the anchor pallet. This in turn results in the wheel having a larger moment of inertia and thus the efficiency of the inhibition decreases. The use of new materials (Si, diamond) has an advantageous effect here, which on the one hand have a significantly lower density than steel and on the other hand, in the case of diamond, a very high modulus of elasticity (700 GPa-1100 GPa) combined with a high bending fracture stress (1 GPa-10GPa). The latter parameters allow the components to be skeletonized even more, thereby reducing the moment of inertia and thus increasing the efficiency of the escapement again. However, the new materials have a significantly lower density than metallic materials and the new processing methods can achieve a higher degree of skeletonization. The density is preferably less than 4.5 g/cm ³ , particularly preferably 1-4 g/cm ³ . Furthermore, the wheel can be designed using a so-called more level technique, whereby the outer contour of the escape wheel, for example, is first etched out of a thicker plate (so-called wafer) using photolithography and reactive ion etching. The advantage here is that the wheel is held in the remaining wafer with small bars so that each component does not have to be manipulated individually. In a further step, the inner region of the gear is thinned using a further photolithography step, which is positioned or adjusted precisely to the first etching step. The second step can also be carried out from the back of the wafer. If necessary, this step can also be done before the outer contour of the escape wheel is exposed. Examples of this technology can be found, among others, at www.Sigatec.ch.

All previous approaches in watchmaking fundamentally pursued the approach of minimizing wear and keeping the moment of inertia of the components as small as possible. That's why the escape wheel was always made as thin as possible. Surprisingly, it has now been found that in the case of hard materials such as diamond or silicon, it is more advantageous to stimulate wear in a targeted manner. According to the invention, it is therefore proposed to invert the arrangement of the escapement in such a way that the armature has the so-called defined geometric contact surface A ₀ . By using materials with low density and high mechanical strength, the escape wheel (or at least the piston teeth of the escape wheel) can now be made thicker than the pallets of the anchor.

According to the present invention, it is provided that the thickness of the pallets is 50 to 180 µm and the thickness of the teeth of the gear is 100 to 500 µm, it being advantageous to make the thickness of the pallets smaller than the thickness of the teeth. According to the invention, the thickness is understood to mean the vertical distance from the top to the bottom of the tooth or the pallet.

According to the invention, as explained above, the flank of the pallets is designed such that whenever there is contact with the front flank of the teeth of the wheel, the front flank of the pallets always engages with the same geometric contact surface A ₀ . According to the invention, the front flanks of the pallets and the front flanks of the teeth, which have a non-verticality, are arranged relative to one another in such a way that a symmetrical non-verticality (clearance angle α) of at least 0.5°, preferably 1°, particularly preferably 2 ° arises. Typically, around 1° per component edge is typical. It is also not necessary that the respective components have the same deviation from the ideal.

With such an arrangement according to the invention, when the escapement is in operation, in the direction of the relative movement, the above-mentioned defined geometric contact surfaces A ₀ in the form of a band, which have a width of 0.5 to 5 μm, arise in the front flanks of the teeth.

These geometric contact surfaces A ₀ obviously arise from the abrasion of the rough component surfaces against each other. It has now been shown that in addition to the geometric contact surface A ₀ , when the escapement runs in, another, namely a polished or partially polished, real contact surface A _R is formed, which is only a partial surface of the geometric contact surface A ₀ . It has been shown that the ratio of the real contact area A _R to the geometric contact area A ₀ is between 20 and 90%.

With regard to the geometric arrangement of the flanks of the pallets, it is preferred if the front flank of the pallet is roof-shaped with a central ridge or cylindrical with an outwardly curved surface (cambered).

According to the invention there is a further embodiment in that the front flank of the pallets is formed in a non-vertical arrangement in relation to the top or bottom of the pallet. In this case, the front flank of the pallets can be designed as a smooth, flat surface which deviates from the vertical in relation to the surface of the pallets by a maximum of t 3°, preferably ± 1°, particularly preferably by less than ± 0.5°.

However, the invention also includes all other embodiments in which the front flank of the pallets is designed in such a way that a defined geometric contact surface A ₀ is created in the front flank of the teeth of the wheel.

In an alternative anchor escapement, which is not part of the invention, the pallets are tilted against the contact surface of the anchor gear, so that not the side end surface of the anchor gear but only the upper edge of the tooth rests on the pallet and transversely to the direction of the edge this surface slides.

It is also preferred if the escape wheel is made of silicon and at least the contact areas of the teeth of the gear also have a hard material coating, preferably made of diamond, similar to the pallets.

In the preferred case, both the anchor pallets as well as the gear and the radially outwardly directed teeth are made of silicon and have a hard material coating.

The hard material coating of the teeth and the pallets has a layer thickness of 1 to 100 µm, preferably 5 to 50 µm and is selected from silicon dioxide, non-stoichiometric oxides with the formula Si _x O _y , where x and y are integers, from silicon oxynitrides or made of silicon carbides, silicon nitride and/or diamond. In the case of silicon components, the contact surfaces can also simply be thermally oxidized (e.g. according to EP 1 904 901 ).

It is preferred if the hard material coating is a coating made of nanocrystalline diamond. In particular, those embodiments in which both the gear and the pallets have a hard material coating made of nanocrystalline diamond are preferred. Coatings that have 96 to 97% sp ³ bound carbon with a grain size of 9 nm are preferred.

It has also been shown that it is advantageous if the nanocrystalline diamond layer has a surface roughness of 3 to 100 nm Rms, preferably 1 to 30 nm Rms, particularly preferably 1-7 nm Rms. The roughness Rms is understood to mean the squared roughness, which corresponds to the root mean square. A nanocrystalline diamond layer with such a low surface roughness requires correspondingly less running-in distance/start-up time, which leads to a shorter running-in phase in order to achieve a minimum and constant coefficient of friction.

It is further preferred if the crystalline domains of the nanocrystalline diamond layer have an average grain size d ₅₀ of 0.5 nm to 50 nm, preferably of 1 nm to 20 nm, particularly preferably of 1 nm to 10 nm. The advantage of such an embodiment is that a nanocrystalline diamond layer that is as homogeneous and uniform as possible and has a very small grain size, as described above, is created. Smaller grains inevitably increase the grain boundary volume. If the grains are small in relation to the surface roughness, the running-in phase can also be accelerated. The connections are in Figure 7 shown. A grain boundary volume of 0 to 50%, preferably 10 to 30%, is preferred. Because separating a single grain from the composite is easier than grinding a large grain smooth. It is therefore advantageous if the crystalline domains are smaller than 0.5xR _t , preferably 0.2xR _t , particularly preferably 0.1xRt of the remaining surface roughness. In this case, however, the absolute surface roughness R _t (roughness depth), measured as a peak-to-valley value, must be taken into account for the roughness R _t . R _t is calculated from the difference between the maximum peak height R _p and the maximum peak depth R _v .

Since both pure silicon and diamond are electrical insulators, it is also proposed to electrically dope both materials so that they have at least a low electrical conductivity. This allows electrostatic charging effects to be avoided. The doping processes for silicon are sufficiently described in the literature. For diamond, it is suggested to use doping with boron or with nitrogen or ammonia. ( N. Wiora et al., Synthesis and Characterization of n-type Nitrogenated, Nanocrystalline Diamond Micron Materials and Nanomaterials, 15:96-98 ). It is also proposed that in the case of insulating materials, the friction partners consist of the same material and therefore have the same work function. This means that electrostatic charging due to mere frictional contact can be largely ruled out.

• Figur 1 zeigt schematisch die Draufsicht auf eine Schweizer Ankerhemmung in den beiden Ruhelagen.
• Figur 2 zeigt den dynamischen Implusionsvorgang einer Schweizer Ankerhemmung wie in Figur 1 dargestellt.
• Figur 3 zeigt nun in vergrößerter Darstellung das Verhalten der Kontaktflächen einer Schweizer Ankerhemmung des Standes der Technik während des Impulsionsvorgangs.
• Figur 4 zeigt eine entsprechende schematische Darstellung bei einer Lageveränderung der Uhr.
• Figur 5 zeigt nun eine erfindungsgemäße Konfiguration der Kontaktflächen wie sie sich während des Impulsionsvorgangs bei der Erfindung darstellen.
• Figur 6 zeigt die erfindungsgemäße Ausführungsform in vergrößerter Darstellung bei einer Lagerveränderung der Uhr.
• Figur 7 zeigt eine grafische Darstellung, wie sich die Korngröße zum Korngrenzvolumen verhält.
• Figur 8 zeigt Rasterelektronenmikroskopaufnahmen eines Kolbenzahnes nach 9 Tagen ununterbrochenen Lauf der Hemmung.
• Figur 9 zeigt die geometrische Kontaktfläche A₀ in vergrößerter Darstellung sowie die reale Kontaktfläche A_R.
• Figur 10 zeigt experimentell ermittelte Amplituden und Gangwerte eines ETA-Kalibertyps 2892A2 bestückt mit einer konventionellen diamantbeschichteten Hemmung.
• Figur 11 zeigt nun zum Vergleich ebenfalls die experimentell ermittelten Amplituden und Gangwerte eines ETA-Kalibertyps 2892A2 bestückt mit einer erfindungsgemäßen diamantbeschichteten Hemmung.
• Figuren 12 bis 14 zeigen weitere erfindungsgemäße Ausführungsformen in schematischer Darstellung, wie sie sich während des Implusionsvorganges bei der Erfindung darstellen.

The invention is described in more detail below merely by way of example using several figures.

• Figure 1 shows schematically the top view of a Swiss anchor escapement in the two rest positions.
• Figure 2 shows the dynamic implusion process of a Swiss anchor escapement as in Figure 1 shown.
• Figure 3 now shows an enlarged view of the behavior of the contact surfaces of a state-of-the-art Swiss anchor escapement during the impulse process.
• Figure 4 shows a corresponding schematic representation when the position of the watch changes.
• Figure 5 now shows a configuration according to the invention of the contact surfaces as they appear during the impulse process in the invention.
• Figure 6 shows the embodiment according to the invention in an enlarged view when the bearing of the watch is changed.
• Figure 7 shows a graphic representation of how the grain size relates to the grain boundary volume.
• Figure 8 shows scanning electron microscope images of a piston tooth after 9 days of continuous running of the escapement.
• Figure 9 shows the geometric contact area A ₀ in an enlarged view as well as the real contact area A _R .
• Figure 10 shows experimentally determined amplitudes and rate values of an ETA caliber type 2892A2 equipped with a conventional diamond-coated escapement.
• Figure 11 now shows for comparison the experimentally determined amplitudes and rate values of an ETA caliber type 2892A2 equipped with a diamond-coated escapement according to the invention.
• Figures 12 to 14 show further embodiments according to the invention in a schematic representation, as they appear during the impulsion process in the invention.

Figure 1 now shows a Swiss anchor escapement in the top view. The Figure 1A shows the state of the anchor 1 and the wheel 4 at the start and the Figure 1B after the unrest has passed zero.

Like from the Figure 1A As can be seen, the escape wheel 4 has 20 piston teeth 5 in the case shown here. The armature 1 has an input pallet 2 and an output pallet 2 ', which alternately engage with the piston teeth 5 of the wheel 4. The Figures 1A and 1B show the rest states in each case. In Figure 1B The new position of the piston tooth 5 is also shown. The stopper 20 serves as a stop for the anchor 4. Alternatively, such a stop can be dispensed with by introducing a shoulder in the anchor pallets. Such an embodiment is, for example, in the Swiss patent specification CH 5 67293 , here in particular the Figures 5 and 6 , described. It is now essential that with a Swiss anchor escapement of the prior art, with each revolution of the wheel 4, all 20 piston teeth 5 of the wheel 4 interact once with both the input pallet 2 and the output pallet 2 '. In the Figure 1 The direction of rotation is indicated by the arrow.

Figure 2 now shows the one in the Figure 1 Swiss anchor escapement described in more detail, here during the dynamic implusion process (interaction between piston tooth 5 and anchor pallet 2 or 2 '). In the enlarged view (Detail A), the essential areas are shown enlarged. The piston tooth 5 of the wheel 4 has already left the rest surface 22 and is located on the lifting surface 23 of the input pallet 2. Due to the applied torque of the gear train, the piston tooth 5 now slides over the lifting surface 23 of the input pallet 2 and thereby pushes the anchor pallet 2 back. For a better understanding of the process, see: Figure 3 referred. The position at which the pallet 2 is in engagement with the anchor 5 is shown at 21.

Figure 3 now shows in an enlarged view how the contact surfaces relate to each other. In order to ensure that the piston teeth 5 of the escape wheel 4 engage securely in the pallets 2, 2' of the anchor 1, the pallets 2, 2' of the anchor 1 are made thicker than the piston teeth 5 of the wheel. This can be seen in particular from section AA. For a better understanding of the course of the impulse process in a Swiss anchor escapement of the prior art, B is shown in more detail, with 25 again showing the point at which the pallet 2 is in engagement with the tooth 5. Due to the manufacturing process, the flanks 12 of the tooth 5 and the flanks 8 of the pallets 2 of the components never have an angle of 90° to the surface (so-called non-verticality). As a consequence, the components can be installed in such a way that the functional surfaces are plane-parallel to one another. The components are now usually mounted in such a way that the flanks are aligned with one another, so that a clearance angle α results. In the case shown here, a symmetrical non-verticality of 2° was assumed. Around 1° per component edge is typical. It is now not necessary that the functional surfaces of wheel 4 and armature 1 have to have the same deviation from the ideal.

Figure 4 now shows the behavior of the components to one another when the position of the clock changes. In Figure 4 The starting position is shown in the left part at section AA and detail C, as previously shown in Figure 3 has been described in more detail. If the position of the watch changes, the bearing play also changes, ie the relative position of the components to one another. This results in a shift Δh between the piston tooth 5 and the input pallet 2, as shown in detail B in the right half of the Figure 4 is shown. It is important that the piston tooth 5 of the wheel 4 is now engaged at point 25 due to the clearance angle α, but still works on the same contact surface A ₀ . The contact point 25 of the lever surface 23 of the input side 2, however, is now shifted by the amount Δh.

Figure 5 now shows an embodiment according to the invention. Analogous to the embodiment of the prior art described above, the flanks 12 of the piston tooth 5 as well as the flank 8 of the pallet 2 have a non-verticality due to the manufacturing process. In the solution according to the invention, too, the components are mounted in such a way that the flanks 8, 12 are aligned with one another, so that analogous to Figures 3 and 4 a clearance angle α results. This again is in Figure 5 shown in detail B. In the embodiment according to the Figure 5 a symmetrical non-verticality of 2° was again assumed. In the invention, too, about 1° per component flank 8 or 12 is preferred. It is also not necessary that the functional surfaces of wheel 4 and armature 1 have to have the same deviation from the ideal. However, what differs from the prior art is that the piston tooth 5 of the wheel was made thicker than the thickness of the pallets 2 of the anchor 4. This also ensures that the piston teeth 5 engage securely in the anchor pallet 2. The contact point is again marked 30. The improved function resulting from the arrangement according to the invention results from Figure 6 .

Figure 6 shows the sectional view Figure 5 now if there is a change in the storage of the watch. This occurs again when the clock is turned, for example. The bearing play changes the relative position of the components to one another. This also results in a shift Δh of the contact points 30 between the piston tooth 5 and the input pallet 2. What is essential to the invention is that the lever surface of the pallet 2, due to the clearance angle α, now still works on the same geometric contact surface A ₀ (defined contact surfaces). The contact point 30 of the piston tooth 5 of the escape wheel 4, however, is shifted by the amount Δh.

What is now crucial is that, compared to the prior art, the defined contact surface of the anchor pallet experiences n-fold wear due to this arrangement, so that it can run in more quickly. The inverse arrangement according to the invention thus achieves a significantly improved running-in behavior of the components, so that the clock is ready for operation at an earlier point in time.

Figure 7 now shows how the grain boundary volume relates to the grain size. How out Figure 7 As can be seen, smaller grains inevitably result in an increased grain boundary volume. A grain boundary volume of 0 to 50% is preferred, preferably 10 to 30%.

Figure 8 shows in an enlarged view ( Figure 8a ) a piston tooth 5 and in Figure 8b the in Figure 8a marked enlarged section

The representation in Figure 8b shows a piston tooth 5 according to the invention after 9 days of continuous running of the escapement. The geometric contact area A ₀ , which is represented by the dark part, can have a width of 0.5 to 20 μm, preferably from 1 to 10 μm, most preferably from 1 to 5 μm. The real contact area A _R (light stripe) is limited to the uppermost part of the component due to the non-verticality of the component. Under 25,000x magnification it can be seen that the geometric contact surface A ₀ in the example case is only about 1.5 to 2 µm wide. The dark color surrounding the geometric contact surface A ₀ results from abrasion of the originally microscopically rough component surface. The bright area has been polished by the function of the component and is actually in contact with the pallets (2,2'). According to the invention, this is referred to as the real contact area A _R. The difference in brightness makes it possible to determine the geometric and real contact area and to form the quotient.

This is also done in, among other things Figure 9 shown, which is also back in Figure 9a the geometric contact area A ₀ and in Figure 9b the real contact area A _r is shown. If you now form the ratio A _r to A ₀ , the result in this case is 74.6%. During the run-in phase, the real contact area increases until a wing portion T of more than 50% is typically reached. Similar evaluations were also carried out with another caliber of type ETA 2824A2 (also Pointage 20.3, i.e. same escapement components), which, however, works with a significantly higher torque. This also results in a significantly higher surface pressure. Here too, it was determined that a wing ratio of at least 50% is necessary to complete the run-in phase.

In Figure 10 are now experimentally determined amplitudes and rate values of an ETA caliber type 2892A2 equipped with a conventional diamond-coated escapement with an sp ² -containing top layer according to the European patent EP 2 236 455 B2 shown. This escapement is lubrication-free.

The amplitude values ( Fig. 10a ) are measured using a Witschi M1 timer (acoustically) and checked optically. The values shown are arithmetic mean values from 6 layers each (dial top, bottom; crown top, right, bottom, left). The measurement interval was 30 seconds in each case, the stabilization time was also 30s. With the average rate deviation ( Fig. 10b ) these are also arithmetic mean values from 6 layers each (see amplitude measurement).

The typical drop in amplitude within the first 7 days is clearly visible. The amplitude then slowly increases again and reaches its starting value after approx. 40 days. This behavior is also explained by the shrinkage of the components. The defined contact surface of the wheel is polished by the frictional contact with the anchor pallets. If the airfoil proportion of more than 50% is reached, no further removal takes place and a tribofilm can form on the contact surfaces. The dashed line shows when the run-in phase is complete.

In Figure 11 are those determined experimentally Amplitudes and rate values of an ETA caliber type 2892A2 equipped with the diamond-coated escapement according to the invention without a soft sp2-containing top layer according to. EP 2 236 455 . The escapement runs without lubrication.

The amplitude values ( Fig. 11a ) are in turn measured using a Witschi M1 timer (acoustically) and checked optically. The values shown are again arithmetic mean values from 6 layers each (dial top, bottom; crown top, right, bottom, left). The measuring interval was 30 seconds in each case, the stabilization time was also 30s. With the average rate deviation ( Fig. 11b ) these are also arithmetic mean values from 6 layers each (see amplitude measurement).

It is clear to see that the typical drop in amplitude now takes place within the first 2 days (conventionally around 7). The amplitude then increases very quickly again and exceeds its starting value after about 10 days. This behavior is also explained by the shrinkage of the components. The defined contact surface of the anchor is polished by the frictional contact with the anchor pallets. If the airfoil proportion of more than 50% is reached, no further removal takes place and a tribofilm can form on the contact surfaces. Because the defined contact surface experiences 20 times the wear of a piston tooth, the inlet chamfer is correspondingly shortened. Furthermore, there is a significantly higher stability in the accuracy.

The Figures 12 to 14 now show further embodiments according to the invention how and in what way the piston tooth 5 and the pallet 2 can be designed in order to enable the geometric contact surface A ₀ according to the invention.

In Figure 12a Such an embodiment is shown in plan view and in Figure 12b the in Figure 12a marked cut.

In contrast to the embodiment according to the Figure 5 is now in the embodiment according to Figure 12 the pallet 2 is beveled twice with respect to its flank, so that a defined contact point 30 is created when it engages with the tooth 5.

In Figure 13 Another embodiment is shown, here an embodiment in which the pallet 2 has a rounded edge. In Figure 13b the top view of such a configuration is again shown schematically in detail and in the Figures 13a and 13b in each case an enlarged view of the contact point 30. In Figure 14 A further embodiment is now shown, in which the pallet 2 has a design as already shown in the Figure 12 has been described, but here the piston tooth 5 is now angled. In the Figure 14 A section of the top view of the configuration is shown in A and in the Figures 14b and 14c enlarged representations in each case, whereby the contact point 30 can also be seen here, so that a geometric contact surface A ₀ can then form here again.

Claims (9)

1. Escapement for a mechanical time-measuring device comprising pivotable pallets (1) with pallet-stones (2, 2') and a pallet wheel (4), to which a torque can be applied and which has outwardly directed teeth (5, 5', 5") over its outer circumference, a sliding relative movement being produced during operation of the escapement, in which the front flanks (8) of the pallet-stones (2, 2') are in contact in succession and alternately with the front flanks of the teeth (5, 5', 5"),

   geometric contact faces A₀ in the form of a band being formed in the front flanks (12) of the teeth (5, 5', 5") during operation of the escapement, in the direction of the relative movement, which contact faces have a width of 0.5 to 5 µm, and the front flanks (8) of the pallet-stones (2, 2') being configured such that, during operation of the escapement in the direction of the relative movement, during each contact of one of the front flanks (8) of the pallet-stones (2, 2') with one of the front flanks (12) of the teeth (5, 5', 5") of the wheel (4), the same contact face A₀ of the front flank (8) of the pallet-stones (2, 2') is always engaged with the front flank (12) of the teeth (5, 5', 5") of the wheel (4), wherein at least the flanks (8) of the pallet-stones (2, 2') and the front flank (12) of the teeth (5, 5', 5") have a hard material coating, wherein a polished or slightly polished real contact face A_R is formed in the geometric contact face A₀, wherein the ratio A_R/A₀ is between 20 and 90%,

   characterised in that the hard material coating has a layer thickness of 1 to 100 µm and is selected from silicon oxide, in particular SiO₂, or non-stoichiometric oxides with the formula Si_xO_y, x and y being whole numbers, or from silicon carbides or silicon nitride or diamond, in particular nanocrystalline diamond, or diamond-like carbon (DLC) or ruby or sapphire.
2. Escapement according to claim 1, characterised in that the front flanks (8) of the pallet-stones (2, 2') and the front flanks (12) of the teeth (5, 5', 5") have non-verticality and are disposed relative to each other such that a free angle α of at least 0.1° - 5°, preferably 0.1° - 3° and particularly preferably of 0.1° - 1°, is produced.
3. Escapement according to one of the claims 1 to 2, characterised in that the front flanks (8) of the pallet-stones (2, 2') are configured in the shape of a roof with a central burr or cylindrically with an outwardly curved face, the geometric contact face A₀ being formed by the central burr or by the outwardly curved face.
4. Escapement according to one of the claims 1 to 3, characterised in that the front flanks (8) of the pallet-stones (2, 2') are configured as smooth flat faces which deviate by max. ± 2°, preferably 1°, particularly preferably less than 0.5°, from the vertical, relative to the upper side of the pallet-stones, the geometric contact face A₀ being formed by the outwardly positioned edge.
5. Escapement according to one of the claims 1 to 4,
   characterised in that the thickness of the pallet-stones (2, 2') is 50 to 180 µm and the thickness of the teeth (5, 5', 5") of the toothed wheel (4) is 100 to 250 µm, the thickness of the pallet-stones (2, 2') being less than the thickness of the teeth (5, 5', 5") of the toothed wheel (4), respectively relative to the vertical with respect to the upper side.
6. Escapement according to at least one of the claims 1 to 5, characterised in that the hard material coating has a layer thickness of 5 to 50 µm.
7. An anchor escapement according to one of the preceding claims, characterized in that the nanocrystalline diamond layer has at least one of the following properties:

   a) the nanocrystalline diamond layer has a transverse rupture stress of 1 to 10 GPa, preferably of at least 2 GPa, preferably of at least 5 GPa and particularly preferably at least 7 GPa,

   b) the nanocrystalline diamond layer has a layer thickness in the contact region of 0.5 µm to 100 µm, preferably 2 - 50 µm and particularly preferably 2-10 µm, and

   c) the nanocrystalline diamond layer has a modulus of elasticity of 700 GPa to 1,143 GPa, preferably of 400 GPa to 900 GPa.
8. Escapement according to at least one of the preceding claims, characterised in that at least the escape wheel was manufactured from a material which has a density of 0.5 g/cm3 to 4.5 g/cm3, particularly preferably of 1 to 4 g/cm3.
9. Escapement according to at least one of the claims 1 to 8, characterised in that the pallet-stones (2, 2') and/or the pallets (1) and/or the wheel (4) are manufactured from silicon and are provided with the hard material layer.

Images