Student Abstracts: Engineering at SLAC

Cascading of Ansoft High Frequency Structure Simulator S-matrices. ERIN AYLWARD (Harvard University, Cambridge, MA 02138) VALERY DOLGASHEV (Stanford Linear Accelerator Center, Stanford, CA 94025) .
Design of modern microwave networks and junctions includes extensive use of sophisticated computer simulations. Although some networks or junctions are too complex for direct computer simulation in their entirety, they can be decomposed into simpler subcomponents. Modeling of the entire system can then be reconstructed by cascading S-matrices obtained from the subcomponent simulations. High Frequency Structure Simulator (HFSS), is a commercial program by Ansoft. that can be used to characterize passive microwave devices. It uses the finite element method and advanced techniques such as automatic adaptive mesh generation and refinement to calculate S-parameters and full-wave fields for arbitrarily shaped 3D passive structures. Subcomponent models of waveguide pieces containing inductive irises and also subcomponent models of cells of linear accelerator structures were modeled in HFSS. The software calculated their S-parameters. Then the S-matrix of a larger network built form these subcomponent S-matrices was computed. This was accomplished by performing scattering matrix cascading on the subcomponent S-matrices. It was found that the S-matrix of a network, which was virtually built from coupled subcomponents modeled by HFSS, could be accurately calculated this way and was representative of the entire network.

Theoretical and computer study of electrons (both individual and in bunches) interacting with both static and dynamic electromagnetic fields. JOHN CASTRO II (Oklahoma State University, Stillwater, OK 74075) ROMAN (Stanford Linear Accelerator Center, Stanford, CA 94025) .
The purpose behind our project is to study how the energy of an electron is modulated by interactions with electric and magnetic fields, both static and dynamic. The dynamic fields that we will be concerned with are radiation fields. This is done to see if practical techniques can be developed to generate electron pulses that are in the atto-second range. The reason this can be explored now is because of the availability of low-emittance electron beams from laser driven photocathode RF guns. This technology has been developed to a sufficiently high level only within the last few years. The approach we are studying will be applicable to electron beams in linear accelerators rather than storage rings. We will simulate both the electron beams and the fields they interact with in mathmatical form using the FORTRAN programming language. We will use the Lorenz Force law to describe and simulate the effects that these fields will have on an electron bunch and determine if the effects will allow the compression of sub-intervals of the electron bunch to atto-second lengths. So far, we have developed a program that defines the initial conditions for each electron in an electron pulse. These initial conditions define the coordinates of the bunch in real and momentum space. The program in its finished form will be able to completely simulate the interactions between the electrons and the fields to see if new designs for electron bunch compression are possible. If so, such designs might be implemented into future linear accelerators here at SLAC.

Phase Stability of the Main Drive Line at Stanford Linear . BENJAMIN COTTS (University of Portland, Portland, OR 97203) RON AKRE (Stanford Linear Accelerator Center, Stanford, CA 94025) .
The Linac Coherent Light Source (LCLS) project at SLAC has higher RF phase stability requirements than the presently running system. Currently there is no way to directly measure the RF stability of the Main Drive Line (MDL) to the desired precision. There is a point at each sector of 8 klystron stations, called a head-tail monitor where phase measurements are taken. This point is the intersection of two signals, which both originated on the MDL and then took different paths to the same place. Though part of one of these paths is on the MDL, the phase measurements include more variation than is on the MDL alone. In order to determine the phase stability of the MDL, it was necessary to build an interferometer. Because of time limitations, data extraction was not possible. The design and testing of the interferometer, and predicted results are discussed.

Temperature Control of Beam Line Mirrors. LINDSAY HOPKINS (Spelman College, Atlanta, GA 30314) JOHN BAGNASCO (Stanford Linear Accelerator Center, Stanford, CA 94025) .
Researchers have noticed that the beam position had been drifting vertically, causing it to miss their research samples. It is believed that this is caused by temperature changes in the mirror water-cooling system. These changes cause the mirrors to pitch, moving the position of the beam. This is due to the differences in silicon and copper thermal expansion coefficients. To alleviate this problem, the group will install temperature-controlling equipment to maintain the water temperature to within 0.01°C from the operating temperature of the water-cooling system of 30°C, which would be an improvement from the current fluctuations of ± 0.2°C. The two choices are water mixing valve temperature control system and a direct heating temperature control system. The group decided to use the water mixing valve temperature control system because it is more accurate. This system should allow the beam line group to regulate the water-cooling temperature within the desired range. The system would monitor whether or not the current temperature is at the desired level. Based on the result, the cooling system will mix hot water into the water flow. The cooling system reads voltages so it does not recognize the temperature readings from the temperature detectors. This leads to the need of a medium to translate the temperature readings to voltages that can be understood by the cooling system. My project is to create the medium that will be installed into the system. Hopefully, this should end the instability problem with the water-cooling system of the apparatus.

Portable System for Calibrating Power Losses in NLCTA Components. CATHERINE KEALHOFER (Princeton, Princeton, NJ 08544) JOSEF FRISCH (Stanford Linear Accelerator Center, Stanford, CA 94025) .
A simple technique for monitoring electric fields in the accelerator structures of the NLCTA (Next Linear Collider Test Accelerator) involves picking off some of the microwave power sent to these structures and measuring it. In this context, the calibration of power losses in the relevant components is an important problem. This paper describes the use of a Gunn oscillator in a portable calibration system. Measurements of the oscillator's frequency and amplitude variations with temperature and operating voltage are also presented. In addition, upper limits placed on the oscillator's phase noise indicate other potential applications for these oscillators, for instance in measuring the phase of the RF sent to the accelerator structures.