BE/ABP-SU

BE/ABP-SU/

Shimming test results for V-shaped support of Epucret girder DB02E

Metrology lab, 31st of may,2012

Sylvain GRIFFET, 13/06/2012

EDMS n° 1221024


BE/ABP-SU/

Introduction

Fiducialisation

TestsInitial measurements1st test: shimming

2nd test: repeatability

Conclusion

Index

2/15Sylvain GRIFFET, 13/06/2012

BE/ABP-SU/

After machining of V-supports which equip the Epucret girder DB02E, the dimensional control found location defaults 10 times higher than the tolerances. The V-support adjustments were planned by the design, so we wanted to test the performance of this shimming by a test in a CMM. We chose to test the shimming on the 27 V-support because it only ⌀supports beam tube: no critical tolerance.

All the results are available in a report and in an Excel file (EDMS n° 1221024).

Introduction



BE/ABP-SU/ Fiducialisation



Dimensional control and fiducialisation of the Epucret girder DB02E in the Olivetti CMM (± 6 µm at 3σ), last April (EDMS n°1209791). Coordinate system based on reference planes.

BE/ABP-SU/ Fiducialisation



The tolerances about the location of V-support centres were ±0.005 mm.Location defaults are mainly along vertical direction (up to 49 µm).

Points X (mm) Y (mm) Z (mm)

V-support centresTheoretical values:

X = 75.000, Z=210.000

R49.5-1 75.010 1603.000 209.951R60-1 75.010 1504.002 209.967R60-2 75.010 1222.002 209.964R60-3 75.007 976.002 209.951

R49.5-2 75.001 599.001 209.950R60-4 75.003 534.001 209.968R60-5 74.979 232.001 209.982R13.5 74.980 95.001 210.005

1.5’’ spheres

DBO2E1 -23.208 1799.969 -222.163DBO2E2 -23.171 1575.961 -119.453DBO2E3 -23.226 1353.236 -273.975DBO2E4 -23.104 1109.210 -118.758DBO2E5 -23.047 809.191 -279.663DBO2E6 -23.054 615.229 -126.771DBO2E7 -22.936 370.714 -278.866DBO2E8 34.904 165.117 25.425DBO2E9 -22.904 91.525 -278.014

BE/ABP-SU/ Tests Initial measurements

• Session 0Locations determined by metrological measurements are analysed by “best fit” calculations performed under Spatial Analyser (change of coordinate system is based on the 3 fiducials).

Obtained results are comparable to fiducialisation measurements (within the CMM tolerances of ±6 µm).



Session 0 in April 2012 system Session 0 – April 2012Points X (mm) Y (mm) Z (mm) dX (mm) dY (mm) dZ (mm)

DBO2E7 -22.936 370.714 -278.866 0.001 0.003 0.004DBO2E8 34.904 165.117 25.425 -0.001 0.002 -0.004DBO2E9 -22.904 91.525 -278.014 0.000 -0.006 0.001

R13.5 74.98 95.001 210.005 -0.004 1.519 -0.009

BE/ABP-SU/ Tests1st test : shimming

In the following, the presented results come from “best fit” calculations to allow to compare results to Session 0.• Session 1 : Shimming of V-support:

• -0.05 mm along X (laterally to girder, add of a shim),• + 0.05 mm along Z (vertically, add of a shim).

Good results, just a lateral default less than -0.01 mm.



Session 1 in Session 0 system Session 1 – Session 0Points X (mm) Y (mm) Z (mm) dX (mm) dY (mm) dZ (mm)

DBO2E7 -22.9353 370.7173 -278.862 -0.001 0.001 0.002DBO2E8 34.9032 165.1194 25.4209 0.000 0.000 0.000DBO2E9 -22.9039 91.5193 -278.014 -0.001 0.000 0.002

R13.5 74.9762 96.5198 209.9964 -0.057 0.001 0.051


• Session 2 : The 2 shims were removed (return to initial status of Session 0) OK.

• Session 3 : Add of the 2 shims (return to status of Session 1).

No problem along vertical direction but the lateral default is about -0.035 mm along X.




DBO2E7 -22.9353 370.7173 -278.862 -0.002 0.001 0.002DBO2E8 34.9032 165.1194 25.4209 0.000 -0.001 0.002DBO2E9 -22.9039 91.5193 -278.014 -0.001 0.000 0.002

R13.5 74.9762 96.5198 209.9964 -0.009 0.002 0.005


DBO2E7 -22.9353 370.7173 -278.862 -0.001 0.000 0.002DBO2E8 34.9032 165.1194 25.4209 0.000 -0.001 0.001DBO2E9 -22.9039 91.5193 -278.014 -0.001 0.000 0.003

R13.5 74.9762 96.5198 209.9964 -0.086 0.001 0.050


How this kind of default can appear?

Instead of vertical fixation, the stress applied by the lateral tightening could move the V-support centre (lever arm effect).

We tried to control the torque in the following tests.



BE/ABP-SU/ Tests1st test : Repeatability

• Session 4 : This session of measurements is just a control after a new positioning of the girder in the CMM: the shimming configuration is always the same (like for Session 3). Results are comparable to Session 3.

In next sessions, the presented results come directly from CMM (fiducials are not measured again).




DBO2E7 -22.9353 370.7173 -278.862 0.000 0.000 0.001DBO2E8 34.9032 165.1194 25.4209 0.000 0.000 -0.002DBO2E9 -22.9039 91.5193 -278.014 0.000 -0.001 0.001

R13.5 74.9762 96.5198 209.9964 -0.091 0.021 0.048


• Session 5 : All screws were unscrewed and the side screw was tightened with a torque of 4 N.m (with a dynamometric key).

The default decreased a little along X but it is always about -0.025 mm. Vertically the default reach the CMM tolerance limit.• Session 6 : Following the same procedure, the torque is now 6 N.m.

The transversal default increased but not along Z.



Session 5 in Session 0 system Session 5 – Session 0Points X (mm) Y (mm) Z (mm) dX (mm) dY (mm) dZ (mm)R13.5 74.904 210.047 -0.077 0.059



• Session 7 : After screwing with a torque of 8 N.m the default along X increased again and it is about – 0.045 now.

The solution of the lateral shimming, with a lever arm that generates a displacement of the V-support centre, definitely does not seem to be a reliable solution.

In next sessions, is presented an another procedure tested to try to remove the lever arm effects.





• Session 8 : We tried to maintain the V-support to the lateral contact area by a clamp. The clamp is removed after the tightening of upper screws. In the figure, the applied stress is symbolized by the red arrow: no lever arm now.




The result is very different from previous: the default along X is about what we can expect without any shim! This value proves that with the side screw, the support was not properly pressed against the reference surface.

We checked the repeatability of this technique by a new operation.


• Session 9 : Like for Session 8, we used a clamp and followed the same procedure.

The vertical default seems to be repeatable. But the default along X increased about 0.015 mm: we concluded that this technique doesn’t allow a repeatable positioning of the V-support.



Session 9 in Session 0 system Session 9 – Session 0Points X (mm) Y (mm) Z (mm) dX (mm) dY (mm) dZ (mm)R13.5 74.991 210.043 0.010 0.055

BE/ABP-SU/ Conclusion

The shimming along the vertical direction provides expected and repeatable results. Laterally, the tightening of the screw occasions defaults and a lake of repeatability. This phenomena is due to a lever arm effect. On the girder, the defaults were mainly along the vertical direction.Several possibilities are possible : • Try to adjust all the V-supports along the vertical direction risk of

damaging the lateral adjustment and at least 5 days of work (cf. D. Pugnat).

• Try to machine again the V-supports on the new AP machine fiducialisation and dimensional will have to be performed again.

• We can consider that the location defaults are acceptable for a mock-up Ok for PETS but on this girder, it was decided to align 2 BPM on 99 V-⌀Supports (the vertical default on these 2 V-supports is about 50 µm).



BE/ABP-SU

Documents

shimming of v

vsupport adjustments

location of v

beabpsutests1st test

presented results

good results

beabpsuafter machining

cmm tolerances