Improving Optics Control and Measurement at RHIC

TL;DR

This work targets maximizing RHIC luminosity by ensuring the collision region aligns with the optics minimum $s^*$, despite sizable horizontal beta-beat around $20\%$ and $s^*$ fluctuations. It introduces a linear sensitivity matrix $\boldsymbol{B}$ to move $s^*$ (notably $s^*_x$) via current adjustments while constraining hardware limits, thereby reducing beta-beat by about $10\%$ and stabilizing $s^*$ measurements. To quantify optics, the paper compares model-dependent methods (curve-fit, MAD-X matching) with model-independent one-turn map approaches using OLS and TLS, plus an error analysis that shows TLS excels in high-$\beta$ regions while OLS remains robust in low-$\beta$ regions. Experimental results at IR8 demonstrate that TLS, CF, and HA reduce $\beta^*$ beat and $s^*$ differences relative to R-OP, enabling more consistent optics control; future directions include Bayesian Optimization and additional independent methods to further improve luminosity optimization.

Abstract

In order to aid in luminosity maximization at the interaction point (IP), the collision location $s_{IP}$ must be equivalent to the location of the minimum value of the beta function $s^*$. Accurate optics measurements and $s^*$ movements are therefore essential to luminosity optimization. However, according to current Relativistic Heavy Ion Collider (RHIC) operations measurements, average horizontal beta beat measurements between operating IPs are around $20\%$ along with significant variation in $s^*$ measurements. A sensitivity matrix was shown to successfully move $s^*$ and the linear optics to their desired values using power supply currents at the 8 o'clock interaction region (IR8). A method to measure the linear optics using the one-turn map within IRs was explored and compared with other mature methods. An error analysis was also included for all optics measurements methods. Through these methods, a $10\%$ beat reduction was consistently achieved while moving $s^*_x$ as well significant improvement to variations in $s^*$ measurements. These methods used at the RHIC control room will be updated for future linear optics analysis and control.

Figures (8)

• Figure 1: Measured horizontal (blue) and vertical (red) beta beat from curve fit method vs BPM number. The solid black lines indicate beta beat $\pm .2$. Error bars were calculated to be around $5\%$ and $3\%$ for horizontal and vertical axis respectively.
• Figure 2: Measured horizontal (blue) and vertical (red) beta beat and $s^*$ measurements at IP8 vs BPM number on the left and right respectively. Six TBT datasets with no movement of $s^*$ ($\Delta s^* = 0$) were analyzed. Error bars were determined using 500 Monte Carlo iterations propagated from beta error values in Fig. \ref{['fig:pictures/RHIC_beat']}.
• Figure 3: TBT data from BPM upstream of IP6 for 1024 turns. The top plot shows the horizontal axis and the bottom plot shows the vertical axis.
• Figure 4: Simulated beta measurements from OLS (red) and TLS (blue) vs BPM error level from 5 to 50$\mu m$. The solid black lines indicate model beta at BPM g5-bx (top) and BPM g9-bx (bottom). Error bars were calculated using a Monte Carlo simulation of 500 iterations for each noise level.
• Figure 5: Simulated beta measurements from OLS (red) and TLS (blue) vs phase at g5-bx, zoomed in at $\phi = [-4.5, -.5]$. The solid black line indicate model beta at BPM g5-bx. The dashed black lines indicate the model $\beta$ and the model phase at the g5-bx BPM