The integration of the slim RF window could improve the situation of the RFQ in multiple ways:

• The installation would require much less space on the test stand.
• The RFQ can be stored under vacuum with the RF window in place.
• One could move the RFQ in and out as required.
• The RFQ does not need to be conditioned on the beamline.
• Realignement in the tunnel (within tight limits) does not require breaking the vacuum.

For this reason, mechanical studies on the RFQ with the weight of the ridged wave-guide, the RF window, and a person loading and unloading the RF window in a critical location have been undertaken by Jorge Guardia.

The simulations have been done with a simplified geometry with correct loading conditions, supported in 3 support points as in reality, and with representative elastoplastic material properties.

Jorge showed the strong geometrical deformation in the model before a realingment.

The relevant outcome is that after realignment of the deformed RFQ, 3 principal types of deformation remain:

• Deformation of the quadrupolar shape in the center with changes of up to 6μm between vanes
• A sag along the structure with a sharp change from -20μm/m to +20μm/m in the center
• A torsion varying all along the RFQ within 0.3mrad with a variation in the center of 0.6mrad/m

Stresses slightly higher than the elastic limit of annealed OFE copper (>8 MPa) are found in several areas, leading to locally plastically-deformed areas which are considered to be uncritical.

Jorge also analysed a number of possible measures to reduce the deformation:

• The best solution to reduce deformation in the quadrupole shape would be hollow stiffening profiles that are tightly connected between RF waveguide flanges. The deformation could be reduced to 2µm between vanes.
• Simpler stiffeners between flanges would also help and be easier to install.

Suitbert proposed an alternative solution by changing the location of the supports: There would be one support at each end on the beamline or to one side and a third in the longitudinal center under the ridged waveguide and the window. The following remarks have been collected:

• There would still be torsion on the RFQ.
• The forces on the brazing of the RF flange could be critical as some fraction of the weight of the RFQ would be acting there.
• The central foot could interfere with the water distribution circuits.
• Exchange of the RFQ will require more time in order to make the environment compatible (changing the location of supports, water hoses and anything attached to the supports).

In the discussion, the following points have been raised:

• The deformation is considered to be uncritical with respect to RF field distribution if the RFQ is tuned with the RF window installed. (Alexej Grudiev)
• The beam dynamics effects need to be carefully ananlysed. The deformation is not as small as that it could be neglected on the basis of earlier simulation results. Action: Alessandra Lombardi & Giulia Bellodi
• Pumping of the RFQ window through the ridged wave-guide shall be checked with the vacuum group. Action: Suitbert Ramberger
• Transport and lifting could be critical. The maximum acceleration and the corresponding deformation shall be checked by MME colleagues. Action: Benoît Riffaud & Jorge Guardia
• The tolerances achieved on the RFQ1 manufacturing shall be checked with the CERN workshops. Action: Suitbert Ramberger -> After brazing, all the vane tips were within the tolerance: +/- 20 microns.
• The solutions found for RFQ2 cannot be extrapolated to RFQ3 before the design is completed. Other coupler solutions could be integrated in the RFQ3 design. (Alexej Grudiev)

In order to take the next decisions and to launch the manufacturing of slim RF windows in time, it would be good to have feedback on all these points by the Project Day on 8 December.