Hydrogen embrittlement bake and route advisor

It depends which standard your print cites, and they disagree. Take a Grade 8 cap screw specified at 39 HRC maximum, about 1034 MPa. Under ASTM F1941 the mandatory trigger in 6.4.1 is 40 HRC, so a bake is not mandatory, but 39 HRC is also above the 34 HRC ceiling of the 5.1.4 Note 2 low susceptibility statement, so the standard neither requires a bake nor calls the part safe. Under ASTM B633-19 the threshold is 1200 MPa, so no bake is mandatory. Under the pre-2019 editions of B633 the threshold was 1000 MPa, so the same part does require one. Same print, same part, opposite answers depending on the edition cited.

Because the thresholds sit close enough together that a converted value can land on the wrong side of a boundary. ISO 4042 tiers at 360 HV and 390 HV, and published conversion tables assign more than one Vickers value to a single Rockwell C number. If you enter 39 HRC, ISO 898-1 anchors put that at both 380 HV and 385 HV, either side of nothing but within a few points of the 390 HV tier boundary. So the tool gives a definite verdict only from the standard whose native unit you actually typed, and shows the others as a band with the thresholds named. Selecting an ISO property class is the exception, because ISO 4042 Table 3 maps class to tier directly with no conversion at all.

This page deliberately does not print one. ASTM B850 Table 1 is a list of embrittlement relief classes from which the purchaser specifies a treatment, and B850 4.4 states plainly that the electroplater is not normally in possession of the design and stress information needed to select it. What is safe to state is the ASTM F1941 6.4.2 envelope: 8 hours is typical, and 2 to 24 hours at 350 to 450 degrees Fahrenheit is permitted by agreement. The same clause says it is not considered possible to give an exact baking duration. Any schedule must stay above 130 degrees Celsius, below the tempering temperature minus 50 degrees, and below the coating melting point.

ISO 4042 Annex B.4 says that for zinc-plated fasteners a bake of four hours or less below 200 degrees Celsius is generally inadequate, especially at high coating thickness, and can even be detrimental. The same clause carries its own qualifiers: depending on the type, size, hardness, coating system, coating thickness, coating process, coating permeability and baking temperature, shorter durations are sometimes applied successfully. It also states that the four hour figure is subjective, must not be used as a lot acceptance criterion, and must not be used to assign root cause after a failure.

It removes the electrolytic hydrogen source, which is the main one, but it does not make a part immune. ASTM B695 X3.1 supports the position that mechanical deposition itself introduces no hydrogen charging, because there is no plating current and therefore no cathodic hydrogen. Acidic cleaning or pickling ahead of the process can still charge hydrogen, so the cleaning route matters as much as the deposition route. B695 also carries its own requirements: 6.2.1 calls for a pre-plate stress relief at 190 plus or minus 15 degrees Celsius for at least 3 hours where tensile strength is at or above 1000 MPa with machining, grinding, straightening or cold forming stresses, and 6.6 requires a 48 hour ambient hold before flexural parts are released.

Within 4 hours of plating and preferably within 1 hour, under ASTM F1941 6.4.2.2, with ASTM B633-19 6.5.3 to the same effect. Sequence matters for a second reason: chromate is damaged above about 150 degrees Fahrenheit, so a part baked after chromating can look perfect and fail salt spray. Allow 24 hours of ambient ageing before salt spray testing, per F1941 9.3.