A teardown of GM’s troubled 6.2-liter V8 found a bent pushrod, and the proposed fix is untested

An engine builder who tore down a failed General Motors 6.2-liter L87 V8 found a bent pushrod, a damaged lifter and a damaged camshaft lobe, and now proposes a software change to the engine’s variable oil pump as a fix. The proposal is untested. The teardown was published as a video by the YouTube channel Dave’s Auto Center and summarized on Oct. 16 by Autoblog Senior Editor Jacob Oliva.

AI, explained every day: The Deep View is the go-to daily newsletter for 600,000+ founders, knowledge workers and professionals who want to stay up to date with artificial intelligence.

Get The Deep View →

We may earn a commission if you sign up.

Those findings land on an engine that federal regulators are already scrutinizing on two fronts, and the shop’s theory points somewhere other than the part GM has blamed. The distance between the two accounts is the real story.

What the teardown found inside one L87

According to Autoblog’s write-up, the builder, who is not named in the article, found the bent pushrod and worn lifter alongside the damaged camshaft lobe. He believes the lifter rotated 180 degrees in its guide, probably because plastic retainers failed to hold it in place. The damaged lifter may then have sent metal debris through the engine, which could explain damage to the number-one main bearing, and several connecting-rod bearings show polishing that he reads as metal-on-metal contact.

His conclusion is hedged. The real culprit, as Autoblog puts it, might be a messy combination of crankshaft surface problems and the variable-displacement oil pump. The teardown of a single engine cannot establish whether the oiling problem or the lifter damage came first, or whether either explains every L87 failure.

The untested fix: reprogramming the oil pump

The proposed remedy rests on the builder’s view that the pump lowers oil pressure at idle and low load to save fuel, leaving bearings unprotected. His idea is to reprogram the pump to hold pressure higher in those conditions. He plans to run modified engines on a dynamometer, then disassemble them to check bearing wear and metal debris, and is also considering stronger lifter retainers.

Autoblog describes the approach as experimental and aimed at owners whose engines do not qualify for replacement under GM’s current diagnostic rules. The article does not include a response from GM, and nothing in it suggests the company has evaluated the idea. Until the dynamometer results exist, the proposal is a hypothesis built on one dismantled engine, offered to people who have few other options.

Where federal regulators stand on the L87

NHTSA’s Office of Defects Investigation has followed the engine for nearly two years. It opened preliminary evaluation PE25001 on Jan. 16, 2025, and closed it on Oct. 23, 2025 after GM issued recall 25V-274 on April 24, 2025. The closing resume covers 866,625 vehicles with L87 engines built between March 1, 2021, and May 31, 2024, across the Silverado 1500, Tahoe, Suburban, Sierra 1500, Yukon, Yukon XL and Cadillac Escalade lines. It identifies the defect as connecting rod bearing failure that can cause loss of motive power and raise crash risk, and records GM attributing it to supplier manufacturing and quality issues. The remedy is an inspection followed by either an oil change to a different-viscosity oil or an engine replacement. The agency had received 1,157 allegations of engine bearing failure by that point.

The same day it closed the evaluation, the agency opened engineering analysis EA25007 to examine failures outside the recall’s scope. That record lists 286,051 vehicles, repeats the 1,157 reports of bearing failure and counts 173 of them as involving vehicles outside the recall.

A third file questions the cure itself. Recall query RQ26001, opened Jan. 16, 2026, asks whether the 25V-274 remedy is adequate. It lists 597,571 vehicles and 36 owner questionnaires alleging engine failure after the recall work was done, and in every reported case the remedy had been completed before the failure. Complaints cover both remedy types, the higher-viscosity oil change and the engine replacement. The record lists no crashes, fires, injuries or fatalities.

What the pushrod finding does and does not change

Set side by side, the paper trail and the teardown disagree about emphasis. NHTSA’s file records a bearing problem that GM attributes to supplier manufacturing quality, while the shop builder sees an oiling and valvetrain chain in which a rotated lifter and bent pushrod may be cause or effect. Both could be partly right, since the closing resume speaks only to bearings and the shop looked at a single engine.

For owners, the practical stakes are narrow. The recall remedy remains the route GM and NHTSA recognize, and the open recall query is the place where any pattern of post-remedy failures would be assessed. The shop’s oil pump change is not an approved repair, and Autoblog frames it for people locked out of replacement.

Coverage of related legal action and a separate probe appears in Autoblog’s reporting on multiple lawsuits merging into one case, though the article does not state case details. GM’s next-generation 5.7-liter L76 and 6.6-liter L78 small-block V8s arrive for 2027 and keep cylinder deactivation, according to Autoblog.

This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.

AI, explained every day: Get The Deep View →

We may earn a commission if you sign up.