EllipticalMirrorParam(OE) parameters setting

[kklmn]

EllipticalMirrorParam(OE) parameters setting

[kklmn]

Yes, you're right. If the optical axis is deflected, the parametric representation needs a shifted pitch angle. In the present implementation this is not available. I will implement it in about a week.
Thank you for discovering the problem!

[kklmn]

Hi,
Please try the last update. Please look at the docs and a new test script test_param_mirror.py for a usage example and get back with your results/comments.

[wholingfei]

Hi,
In test script test_param_mirror.py, I changed the inclination to -2.5e-3, something weird happened. Dose that mean the inclination couldn't be negative?

[kklmn]

Yes, you're right. Thanks for finding it!
I've fixed it. Please try the update.

[wholingfei]

Hi, the test script test_param_mirror.py could run correctly in my computer now.
However, I still can't successfully apply the new EllipticalMirrorParam to my own codes. Here is the main building of the optical elements .

    beamLine.TorM = roe.ToroidMirror(beamLine, 'ToroidMirrir', [0., 41000., 0.], 
                                     pitch=2.5e-3, 
                                     roll=-np.pi/2, 
                                     material=None, 
                                     limPhysX=[-20., 20.], limPhysY=[-256., 256.], 
                                     R=[41000., 19000.], r=[41000., 19000.])
    beamLine.Screen2 = rsc.Screen(beamLine, '2ndSource', [0., 60000., 0.])
    beamLine.Screen2.center[0] = beamLine.TorM.center[0] - (beamLine.Screen2.center[1] - 
                                    beamLine.TorM.center[1])*np.tan(2*beamLine.TorM.pitch)
    KB1_center1 = 76440.
    KB1_center0 = beamLine.TorM.center[0] - (KB1_center1- 
                                      beamLine.TorM.center[1])*np.tan(2*beamLine.TorM.pitch)
    beamLine.KB1 = roe.EllipticalMirrorParam(beamLine, 'VKB', [0., 76440, 0.], 
                                             pitch=1.5e-3, material=mPt, 
                                             limPhysY=[-190., 190.], 
                                             p=16440., q=1560., 
                                             isCylindrical=True)

    f1_x = beamLine.TorM.center[0] - (beamLine.Screen2.center[1] - 
                          beamLine.TorM.center[1])*np.tan(2*beamLine.TorM.pitch)
    KB2_center1 = 76870.
    KB2_center0 = beamLine.Screen2.center[0] - (KB2_center1-beamLine.Screen2.center[1]) * 
                                                  np.tan(2*beamLine.TorM.pitch)
    KB2_center2 = beamLine.KB1.center[2] + (KB2_center1-KB1_center1)
                             *np.tan(2*beamLine.KB1.pitch)
    
    #beamLine.KB2 = roe.EllipticalMirrorParam(beamLine, 'HKB', 
    #                                 [KB2_center0, 76870, KB2_center2], 
    #                                 pitch=1.5e-3, material=mPt,  
    #                                 roll=np.pi/2, extraPitch=-5.0e-3, 
    #                                 limPhysY=[-195., 195.], 
    #                                 p=16870., q=1130., isCylindrical=True)
    beamLine.KB2 = roe.EllipticalMirrorParam(beamLine, 'HKB', 
                                   [KB2_center0, 76870, KB2_center2], 
                                   pitch=1.5e-3-5.e-3, material=mPt, 
                                   roll=np.pi/2,
                                   limPhysY=[-195., 195.], 
                                   f1=[f1_x, 60000., 0.], q=1130., 
                                   isCylindrical=True)
    beamLine.KBSampleShort = rsc.Screen(beamLine, 'AtSamplePos', [0., 78000., 0])

The commented out statements are my original treatment. Toroidal mirror provide a second source at Screen2.
I think the main difference between my codes and the test script is that my codes deflect in sagittal plane not the meridian as shown in test script.

I just want to make sure whether it's my own problem in parameter setting or not.
Thanks for your quick update anyway!

[kklmn]

Hi,
I have now implemented the most general case and have tested it with test_param_mirror.py. I have removed pitch from the ellipse calculations and put the general coordinate system transformation. Please try the last update.
Note also that when you use a horizontally deflecting mirror and the incoming beam is not horizontal, you should also define a yaw angle. Specify a sequence for the rotations to get the orientation that you expect. Please look at test_param_mirror.py for yaw.

[wholingfei]

Hi,
it seems that both the test script and my own script could run correctly on my computer now.
Thanks very much for your help!