PulsCar analysis of NHTSA public data, September 2026 Topic: CV axles, CV joints, CV boots and front half shafts License: CC BY 4.0 (attribution: "PulsCar analysis of NHTSA public data, September 2026") ------------------------------------------------------------------ 1. SOURCE FILES ------------------------------------------------------------------ NHTSA public flat files, downloaded 17 September 2026 into a local SQLite database: COMPLAINTS (FLAT_CMPL), RECALLS (FLAT_RCL), TSBS (FLAT_TSB, the manufacturer-communication index), INVESTIGATIONS (FLAT_INV). Field definitions: https://static.nhtsa.gov/odi/ffdd/cmpl/CMPL.txt Complaint rows: 777,040 unique reports received 1 January 2015 to 8 September 2026. ------------------------------------------------------------------ 2. THE COMPLAINT QUERY, AND TWO ROUNDS OF TIGHTENING ------------------------------------------------------------------ Stage 1, FTS5 on the description: "cv axle" OR "cv joint" OR "cv boot" OR "half shaft" OR "halfshaft" OR "axle shaft" OR "drive axle" OR "constant velocity" 1,095 matches. Reading 30 at random showed roughly 40% were off topic: CVT transmissions (owners write "constant velocity transmission"), wheel separations, control arms, and long maintenance histories in which an axle is one line item among ten. Stage 2 required the axle term to sit within 60 characters of a relevant symptom, and excluded CVT language outright: TERM = \bCV\b | C\.V\. | CONSTANT VELOCITY (JOINT|AXLE|BOOT) | HALF ?-?SHAFT | AXLE SHAFT | DRIVE ?SHAFT | DRIVE AXLE SYM = CLICK* | POPPING | CLUNK* | BOOT | GREASE | VIBRAT* | SHUDDER | SNAPPED | TORN | SPLIT | CRACK* | BROKE | BROKEN | SEPARATED | REPLACED | FAILED | FAILURE | NOISE match = TERM within 60 chars of SYM, in either order exclude = \bCVT\b | CONSTANT VELOCITY TRANSMISSION | CONTINUOUSLY VARIABLE Working set: 587 complaints. FALSE-POSITIVE READ. We read 30 randomly selected survivors in full (seed 41). Twenty-eight were on topic. Two were not: a cable that dropped onto a rear drive axle, and a subframe replacement in which a CV axle was one line of a list. Observed false-positive rate: 2 of 30 (7%), down from roughly 40% before tightening. We also tested a symptom-first query — (click OR clicking OR popping) AND (turn OR turning) — which returns 4,101 raw and 2,231 after a proximity regex. Reading a sample showed heavy contamination from starter clicking, steering-column clicking, audio faults and turn signals, so we did not publish any count from it. Every number below comes from the 587-report set. This is a small data set by the standards of this project, and we say so in the article. Model-level counts under about 30 reports should be read as small samples, not as rates. ------------------------------------------------------------------ 3. WHAT WE COMPUTED (587 complaints) ------------------------------------------------------------------ Text shares (regex on the description; every one is a floor, because owners describe what they noticed, not what a technician found): names clicking or popping ............................ 155 (26.4%) names a snapped, broken or separated shaft ........... 139 (23.7%) mentions towing ...................................... 105 (17.9%) names the boot ....................................... 104 (17.7%) names a recall or campaign ........................... 134 (22.8%) names vibration or shudder ............................ 72 (12.3%) mentions warranty ..................................... 74 (12.6%) names grease .......................................... 53 ( 9.0%) describes both sides .................................. 43 ( 7.3%) mentions a parts backorder ............................ 20 ( 3.4%) Flags from the flat file, against the all-complaints baseline of 777,040 reports (5.1% towed, 4.4% crash alleged): towed 7.3% · crash alleged 2.7% · fire 0.9% Mileage: rows with miles <= 0 or >= 400,000 dropped. Only 236 of 587 (40.2%) carry a usable mileage — a low coverage rate that we state in the article. Median 83,947; quartiles 45,000 and 132,000. Makes: Honda 172, Ford 111, Toyota 40, Jeep 33, Ram 21, Chevrolet 20, Nissan 20, Hyundai 18, Tesla 18, Dodge 17. Models: Honda Prologue 115, Ford Maverick Hybrid 21, Honda Accord 20, Honda Fit 19, Ford F-150 15, Ram 1500 15, Ford Explorer 13, Ford Escape 12. Model years: 2024 is the largest single year at 123 reports. THE PROLOGUE CONCENTRATION. 115 of the 587 reports (19.6%) are 2024-2025 Honda Prologue, an electric SUV first sold in 2024. Only three of those reports carry a usable mileage (200, 2,000 and 10,000), so we publish no median for it and describe the mileage evidence as anecdotal. The concentration is real in the file and is corroborated by Honda's own Tech Line article ATS30070 (below); it is not a failure rate, and we do not present it as one. Published files: cv-axle-complaints-by-make.csv top 25 makes cv-axle-top-models.csv top 30 make+model cv-axle-symptom-shares.csv the text shares above cv-axle-bulletins.csv 202 manufacturer communications Bulletin query (FTS on the summary): ("cv joint" OR "cv boot" OR "half shaft" OR halfshaft OR "drive shaft" OR driveshaft) AND (click OR clicking OR boot OR grease OR vibration OR shudder OR noise) 202 distinct NHTSA document IDs, led by Ford 37, Ram 31, Toyota 17, Acura 12, Cadillac 11, Mercedes-Benz 10. ------------------------------------------------------------------ 4. PRIMARY DOCUMENTS OPENED AND READ ------------------------------------------------------------------ Ford Customer Satisfaction Program 25N09, 17 September 2025 - 2022 Maverick, 68,126 US vehicles, no-cost front halfshaft replacement, 10 years or 150,000 miles, transferable (NHTSA ID 11023857) https://static.nhtsa.gov/odi/tsbs/2025/MC-11023857-0001.pdf Honda Tech Line article ATS30070, 19 December 2025 - 2024-2026 Prologue, clicking or ratcheting noise when turning (NHTSA ID 11026973) https://static.nhtsa.gov/odi/tsbs/2026/MC-11026973-0001.pdf Mazda service alert SA-028/26a, 8 July 2026, superseding SA-028/26 - 2024-2026 CX-50, CX-50 Hybrid and CX-30 clicking on turns; chassis-ear procedure; outer ball joint cage machining out of specification (NHTSA ID 11034844) https://static.nhtsa.gov/odi/tsbs/2026/MC-11034844-0001.pdf Toyota T-SB-0028-26, 16 April 2026, superseding T-SB-0082-24 - 2024 Grand Highlander rotational clicking from the front drive shaft on acceleration while turning, "may occur without a grease leak present" (NHTSA ID 11032480) https://static.nhtsa.gov/odi/tsbs/2026/MC-11032480-0001.pdf Lexus L-SB-0033-24 Rev2 - 2023-2024 RX 350/350h/500h and 2024 TX 350/500h, same condition (NHTSA ID 11034816) https://static.nhtsa.gov/odi/tsbs/2026/MC-11034816-0001.pdf Ford SSM 54827 - 2020-2026 Explorer, Police Interceptor Utility and Aviator, front halfshaft inner and outer CV boot kits released for service (NHTSA ID 11032760) https://static.nhtsa.gov/odi/tsbs/2026/MC-11032760-0001.pdf Acura owner notification ONL6YU04092026 (service bulletin 25-005) - 2024 ZDX electric park brake harness rubbing the half-shaft boot; free repair (NHTSA ID 11031503) https://static.nhtsa.gov/odi/tsbs/2026/MC-11031503-0001.pdf ------------------------------------------------------------------ 5. LIMITATIONS ------------------------------------------------------------------ - 587 reports is a small set. Counts are not rates, and they are not weighted by how many vehicles are on the road. - Mileage coverage is only 40.2%, the weakest number in this analysis. - Complaints are self-reported and unverified. - Publicity and warranty disputes drive filings. A model under a known campaign, or one whose parts are on backorder, generates complaints that an equally faulty but quietly repaired model does not. - The bulletin count is an index search on summaries, not a reading of all 202 documents. Only the seven listed in section 4 were opened and read. - www.nhtsa.gov blocks automated requests (HTTP 403); recall records are linked through api.nhtsa.gov, which serves the same NHTSA record. ------------------------------------------------------------------------------ CASE SENSITIVITY — SECTION ADDED 19 SEPTEMBER 2026 ------------------------------------------------------------------------------ NHTSA stores complaint narratives in ALL CAPS for records filed through roughly 2020 and in mixed case for records filed from 2021 on. A case-SENSITIVE search of the narrative field loses about 92% of the complaints filed after 2020, so this is the single largest way an analysis of this database can go wrong while still looking finished. Every query behind this file is case-insensitive. SQLite's LIKE and its FTS5 full-text index both ignore case by default, and no query in this work uses GLOB or any case-sensitive comparison. This was not taken on trust. On 19 September 2026 every published folder was re-checked against the same database, comparing the figures here with what a case-sensitive query would have returned. The result: the complaint set published here (587) is LARGER than the absolute ceiling a case-sensitive search of the same terms could have returned (145 narratives). A case-sensitive query cannot produce a bigger set than itself, so the bug is ruled out for this folder by arithmetic, not by assurance.