AB Presentation 14/10/03

So, if you remember from the meeting two weeks ago, I was having problems simulating the Mylar window of the ionization chamber, in GEANT, the fact that it was just too thin for GEANT to recognize it (see top of page 1 of attachment).
So, I went ahead and made the window 30 times thicker (and 30 times less dense) so that GEANT would recognize the window volume.
And it worked!!
An interactive version of GEANT showing a 1.206MeV loss in energy, of the 14N recoils, through the Mylar window (see bottom of page 1 of attachment).

Next thing was, I used SRIM to calculate the energy loss through both types of Mylar window – the actual window (0.94um thick and a density = 1.39g/cm^3), and the new simulated window (0.00282cm thick and a density = 0.0463g/cm^3). (see page 2 of attachment).
I converted the output from SRIM into Excel, and used the energy data to plot histograms, for both Mylar conditions, and to calculate the mean energy loss (see page 3 of attachment). The actual window gave an energy loss of 1.2247MeV, and the simulated window gave an energy loss of 1.2251MeV.
(SRIM uses more complicated calculations than GEANT, which would explain the 20keV difference. But SRIM finds only a 4keV difference between the two types of Mylar window, which means that the simulated window is good enough to simulate the actual Mylar window).

So, I went ahead with the simulation, adding straggling effects to the simulation, and was pleased to get the double peak in the final energy spectrum. However, the recoils seemed to stop in two positions in the ionization chamber, which I thought was strange (see page 4 of attachment).
Next, I decided to try and explain these to ‘spots’ of recoils, and used SRIM (and a dE/E program, on the Tuda machine) to calculate the stopping distances of 5.2MeV and 6.4MeV, (these energies roughly correspond to the two peak energies on page 4), 14N recoils in isobutane (at the correct density for this run). (see page 5 of the attachment).
SRIM showed that the different energy recoils stop 2cm apart, which corresponded to the two ‘spots’ in GEANT (see page 6 of attachment). The stopping distances are different, but as I said before, SRIM and GEANT are slightly different.

However, recently while running some interactive GEANT simulations, I noticed that the recoils were missing the entrance tube, and passing straight through ionization chamber, into the isobutane (see page 7 of attachment). Peter Gumplinger had recommended to me, that I make the mother volume of the ionization chamber out of vacuum, which is how the recoil has managed to do this. Also, the entrance tube of the ionization chamber, is slightly off centre in the y-direction of the ionization chamber, and because the ionization chamber ‘sits’ in the centre of the beamline, the tube does not.
So, I changed the mother volume to be made of aluminium, and moved the ionization chamber up in the y-direction, so that the tube was in the centre of the beamline. It was at this point in my lab notebook, while doing a scale drawing of the changes that I was making, that I realized that the entrance tube in the simulation was not 5cm in diameter as it should be (see page 8 of attachment). Going back over my work, it was back in early Sept when I was having problems with GEANT, that I was told that in GEANT, all lengths for volumes are given in half values. So I went ahead and halved all the lengths, including the radius’s of the tubes!!!!

So, with the correct ionization chamber measurements, and an aluminium casing around it, I went ahead and run an interactive and batch simulation of GEANT, the interactive version showing that the recoil makes it into the ionization chamber via the tube and Mylar window, and the batch version showing that the recoils only stop as one ‘spot’ (see page 9 of attachment). This is when I realized that the two spots were caused by same passing through the window and some that didn’t make it through the window, and not losing the 1.2MeV through. Of course, SRIM showed that the recoils with 6.4MeV stop 2cm further than 5.2MeV recoils, but the reason there were 6.4MeV was because they hadn’t lost and energy through the window, because it hadn’t passed through it!

And finally, I have also added dE graphs into GEANT, which show the energy of the recoils in each anode of the ionization chamber (see page 10 of attachment).