*********************************************************
	MEBT REBUNCHING CAVITY FREQUENCY STUDIES
*********************************************************

Simulation plan:
*************************************************
To study the effect of using different sized ports:
Model 81: 4 LARGE PORTS (CF63), EMPTY
Model 82: 2 LARGE PORTS, 2 SMALL PORTS (CF40), EMPTY

To study the tuning range of one large (CF63) tuner:
Model 83: Model 81 with one tuner 10mm OUT
Model 84: Model 81 with one tuner 20mm IN
Model 85: Model 81 with one tuner 40mm IN

To study the change of frequency due to plating / vacuum deformation:
Model 86: Model 81 with 50 microns added to the inner surface
Model 87: Model 81 with 100 microns added to the inner surface
Model 88: Model 81 with 200 microns added to the inner surface

Perhaps model a vacuum deformed cavity? Pete can supply .SAT file.

To help JP with GL work:
648 MHz cavity: CAVITY_648.sat
972 MHz cavity: CAVITY_972.sat

Most effective place to tune:
When Morteza has modelled these we need a print-out of the magnetic field distribution so that we can assess where tuning is most effective. This may guide concept of a removable section for skimming.

On cavity frequency change:
Nose separation = 16mm
Vacuum deformation for mild steel cavity = 80 microns (change of nose separation)
0.080mm -> 0.5% change -> capacitance rise of 0.5%
F= sqrt 1/LC -> 0.25% drop in frequency = 0.8 MHz
Equates to 10MHz per mm in nose region.
Does not match Scott’s 2MHz/mm.




Filename explanation:
*************************************************
81_ReBunchingCavity5_R=20mm_D=568mm_P=66x4

ReBunchingCavity5:	it comes from the 5th series of cavities I have designed.
			This just helps me know where the model came from.
R=20mm:			it has a radius of 20mm on its outer edge.
D= 568mm:		the overall diameter (internally) is 568mm.
P=66x4:			it has 4, 66mm diameter ports


File explanation:
*************************************************
This file was created with a smaller internal radius to allow for larger ports.
Why? It was felt by Alan and others that the small (2%) drop in Q caused by the smaller radii was offset by the benefit or quicker pumping (and a broader tuner range).

Result:
*************************************************
81_ReBunchingCavity5_R=20mm_D=568mm_P=66x4.sat
Date: 8th August 2013
CAD modelling: Pete
COMSOL modelling: Morteza
Frequency: 327.5387 MHz
First harmonic frequency: 668.7301 MHz

Conclusion:
*************************************************
Cavity frequency is approx 3.5MHz too high.
Pete looked at 'trial' ANSYS models to help to choose new diameter.
New diameter chosen as 576mm

Result:
*************************************************
81_ReBunchingCavity5_R=20mm_D=576mm_P=66x4.sat
Date: 9th August 2013
CAD modelling: Pete
COMSOL modelling: Morteza
Frequency: 324.490 MHz
First harmonic frequency: 658.876 MHz

Conclusion:
*************************************************
Cavity frequency is approx 0.5MHz too high.
Aim for a couple of hundred kHz below 324MHz.
Use simple straight line graph to of known two diameters and frequencies to pick new diameter.
See "Diameter_prediction.jpeg"
New diameter chosen as 577.5mm

Result:
*************************************************
81_ReBunchingCavity5_R=20mm_D=577.5mm_P=66x4.sat
Date: 9th August 2013
CAD modelling: Pete
COMSOL modelling: Morteza
Frequency: 323.925 MHz
First harmonic frequency: 657.061 MHz

Conclusion:
*************************************************
We have found our baseline model.
JP has requested that future simulations include:
A field pattern to see if quadrupole modes are present.
A check to see whether any quadrupole/dipole modes exist below the fundamental mode.