© 2016 IEEE. The U.S. LHC Accelerator Research Program (LARP) and CERN are combining efforts for the HiLumi-LHC upgrade to design and fabricate 150-mm-aperture interaction region quadrupoles with a nominal gradient of 130 T/m using Nb3Sn. To successfully produce the necessary long MQXF triplets, the HiLumi-LHC collaboration is systematically reducing risk and design modification by heavily relying upon the experience gained from the successful 120-mm-aperture LARP HQ program. First-generation MQXF short (MQXFS) coils were predominately a scaling up of the HQ quadrupole design, allowing comparable cable expansion during Nb3Sn formation heat treatment and increased insulation fraction for electrical robustness. A total of 13 first-generation MQXFS coils were fabricated between LARP and CERN. Systematic differences in coil size, coil alignment symmetry, and coil length contraction during heat treatment are observed and likely due to slight variances in tooling and insulation/cable systems. Analysis of coil cross sections indicate that field-shaping wedges and adjacent coil turns are systematically displaced from the nominal location and the cable is expanding less than nominally designed. A second-generation MQXF coil design seeks to correct the expansion and displacement discrepancies by increasing insulation and adding adjustable shims at the coil pole and midplanes to correct allowed magnetic field harmonics.