Hardness and Distortion After Quench and Temper of a 4140 Shaft
A 4140 through-hardened drive shaft, 70.0 mm finished diameter by 900 mm long, is specified at 285-321 HB with a minimum 12 mm hardened depth. The current route is: rough turn to 72.0 mm, austenitize at 845 C, oil quench, temper at 600 C, then finish grind to size. Incoming inspection of the last lot found: (1) 0.45 mm TIR bend on three of eight shafts, (2) one shaft reading 258 HB at the surface but 300 HB at 10 mm depth on a bench Brinell tester, and (3) a second shaft that read 305 HB on the bench Brinell but 275 "HB equivalent" when an inspector used a portable Leeb tester on the ground journal and converted per a wall chart. Grinding stock is 1.0 mm per side. The shop wants to grind through the problems and ship, and asks you to confirm.
1. Can the bent shafts be salvaged by grinding alone? Compute the stock consumption and state the limit condition.
Read the worked answer
A 0.45 mm TIR bend means the axis of the finished journal must be brought back on center. Grinding removes material from the high side, so eliminating the bend consumes approximately 0.45/2 = 0.225 mm of stock on one side before any size correction.
Remaining stock on the high side: 1.0 - 0.225 = 0.775 mm.
Now subtract what must come off for cleanliness and decarburization. If the surface layer is decarburized (see the 258 HB surface reading), that layer must be fully removed or the finished part carries a soft, low-fatigue-strength skin. Typical decarb depths for an uncontrolled-atmosphere 845 C austenitize run are 0.3-0.6 mm. At 0.5 mm decarb, required removal on the high side is 0.225 + 0.5 = 0.775 mm - exactly the available stock, i.e., zero margin.
The limit condition is therefore: bend/2 + decarb depth + grinding wheel allowance must be less than the per-side stock. Here it is not met with any margin, and grinding 0.775 mm off one side while removing ~0.3 mm off the other creates a strongly unbalanced residual stress pattern that will bow the shaft after grinding and can produce grinding burn and re-hardening cracks on a 300 HB part.
Recommendation: do not grind these as-is. Straighten by press or by hot spot after tempering (before finish grind), then stress relieve at roughly 540 C to remove the straightening residual stress, then grind. Verify decarb depth metallographically (ASTM E1077) on a cut-up sample before committing the lot.
2. Explain the 258 HB surface / 300 HB core gradient and the portable-tester discrepancy. What would you require before accepting any hardness data from this lot?
Read the worked answer
The gradient is a decarburization signature. A through-hardened 4140 section should be uniform to within a few points of hardness across 12 mm. A surface that is 42 points softer than at 10 mm depth means carbon was lost from the surface during austenitization and that skin did not transform to martensite. A grinding-burn soft skin (overtempered martensite) is the other candidate, but burn is shallow, patchy, and accompanied by a re-hardened untempered layer - a microhardness traverse and a nital etch distinguish the two in one section. This is a furnace atmosphere / protective atmosphere failure, not a material failure.
The portable-tester discrepancy is a method problem, not a part problem. Leeb (rebound) hardness is sensitive to surface finish, mass, curvature and rigidity; it is not equivalent to Brinell on a ground journal of a 900 mm shaft, and ASTM E140 conversion tables between scales are approximate and explicitly not to be used for acceptance. The 30-point gap is larger than any real hardness variation in a 4140 section and should be discarded, not averaged into the lot.
Before accepting data, require: bench Brinell (or a calibrated bench tester) at defined locations (surface, 6 mm, 12 mm depth on a sectioned coupon), the conversion basis stated, decarb depth per ASTM E1077, and a nital etch on one part to rule out grinding burn.
3. Recommend a process change to prevent recurrence. Address sequence, atmosphere, quench practice, and where straightening belongs.
Read the worked answer
Sequence. Rough machine with 1.5 mm per side stock (not 1.0), then stress relieve before finish machining (rough out, 650 C stress relief, semi-finish), so that the bulk of the machining residual stress is gone before the hardening step. Finish grind only after heat treat.
Atmosphere. Austenitize in a atmosphere-controlled or vacuum furnace with carbon potential set to the 4140 carbon level (0.40%), or provide at least 1.0 mm of grinding stock dedicated to removing decarb. Verify with a shim/coupon per heat.
Quench and fixturing. Quench vertically with agitation, not on the flat and not in a basket of stacked parts; asymmetric cooling is the dominant cause of the 0.45 mm bow. A fixture or a properly agitated vertical quench reduces bend to a few tenths.
Straightening. Straighten after tempering and while warm, never after quenching (quench cracks) and never after finish grinding. Follow with a 540 C stress relief for at least one hour per 25 mm of section.
Specification. Replace "285-321 HB" with the scale, the test method, the test location and depth, and add a bend limit at the semi-finished stage. Hardness alone does not describe a heat-treated shaft; depth of hardening and straightness do.