|
|
 |
Scientific news from Division 1 |
Adjustment of the lever arm lengths at the 20 kN·m torque standard machine
|
|
The 20 kN·m torque standard machine of PTB ranks among the best facilities of its kind in the world. To ensure this leading position also in the future, investigations and comparisons are performed at regular intervals. In the last few years, a small asymmetry in the arm lengths of the lever (the right arm was too short by approx. 10 µm, the left too long by approx. the same amount) has been found, which has now been trickily corrected.
|
|
When the 20 kN·m torque standard machine was developed and constructed approximately 10 years ago as a lever-mass system with air bearing, it was - for constructional reasons - necessary to design the lateral parts of the lever (approx. 2 m in length) the lever side plates with tapered ends so that the lever in the centre (on the air bearing) became broader than at the ends, where the forces are applied. At that time regarded as a necessary malady, this design now proves to be of advantage - when specific possibilities have been provided.
Close to their tapered parts the lever side plates are connected with rod-shaped stabilizers that are equipped with screw threads. This allows a self-adjusting connection between lever side plate and rod to be created with the aid of nuts and a combination of spherical and conical washers. Via the nuts, the distance of the lever side plates is now also slightly variable. In doing so it turns out that the length of the corresponding lever arm also changes as a result of the tiny deformation of the lever side plates - although to a clearly smaller extent. Calculations using the FEM the Finite Element Method (Figure 1, top): for reasons of symmetry, only half a lever side plate (the quarter lever) had to be modelled; the screwed joint is not shown in the model had been performed in the past. From these calculations it was known that the stabilizers do not have any carrying function, i.e. that they do not have to bear great loads, but basically influence only the mechanical and geometric properties of the complete lever, such as stiffness and length.
The nuts (Figure 1 below) were now turned by an angle of approx. 5° against the rod. In the case of a flank lead of 3 mm, this results in a theoretical distance variation of the lever side plates by 42 µm. The resulting lever arm length variation amounted to approx. 10 µm. To be able to perform these variations also in a controlled way, a very stable torque transducer connected to a high-precision amplifier was installed in the facility, and measurements were carried out after each adjusting step which showed the variation of the lever arm length. This method allowed the asymmetry of both sides to be reduced to between 1 µm and 2 µm. The absolute comparison was performed in the partial range with the 1 kN·m torque standard machine, whose lever arm length had been determined on a coordinate measuring machine with an uncertainty of only 1 µm (at a lever arm length of 500 mm), and which has recently proved its worth as pilot measuring device for torque in the first key comparison (CIPM key comparison CCM.T-K1) worldwide.
|
|
 |
| Figure 1: Results of the FEM calculations at the quarter model of the tapered lever of the 20 kNm torque standard machine (tapered both in horizontal and upward directions) for an applied force F of 20 kN (top left: vertical deflection, colour scale in m, right: horizontal length variation, colour scale in m, X axis of rotation) as well as nut with spherical (convex) and conical (concave) washers on stabilizer (photo below).
|
|
Contact person: |
|
Dirk Röske, FB 1.2, AG 1.22, dirk.roeske@ptb.de
|
© Physikalisch-Technische Bundesanstalt
Page created: 16.07.2007, last update: 16.07.2007,
Dirk Röske
|
|
